Peer Revenue An exchange of ideas

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Computation and Language

cs.CL Watch Computation and Language

Monday 6 July

Computation and Language

The yes-no bias of large language models reflects answer order and wording, not shifts in moral judgment

Haonan Huang

Large language models (LLMs) increasingly issue judgments read as binary verdicts, and a growing literature reports such judgments shifting under logically irrelevant changes of wording - among them an amplified yes-no bias on moral dilemmas, absent in humans. A single framing cannot say what such a shift is: in a yes/no question the word "no" is at once logical verdict, lexical token, and last-printed option. We introduce a psychometric battery that separates these: crossed symmetrization - every logically irrelevant factor flipped in balanced pairs - across a corpus of question forms. A graded rating across logically equivalent forms recovers a coherent internal moral scale: frontier models' stance $\theta$ is nearly format-invariant (cross-form incoherence 0.12-0.21 on a $\pm 1$ axis); small open-weight models fail in model-specific ways. Forcing the verdict through yes/no overlays a decomposable artifact: an order bias toward the last-printed option - opposite to classic human primacy - plus a lexical pull toward the word "no"; the artifact is substantial only in the Claude models (story-averaged -0.32 to -0.86), $\approx 0$ for GPT-5.5 and Gemini, and shrinks under extended reasoning. The word and the verdict share one token; swapping the words for arbitrary labels separates them, and the verdict-attached logical bias proves $\approx 0$ for every frontier model, while model-specific label and order attachments remain: the models are not drawn toward rejecting - the pull follows the printed surface, not the verdict it carries. A minimal model, $P = \sigma((\theta \pm m)/s)$, summarizes any such artifact by a framing susceptibility m and a moral decisiveness s, measurably distinct from sampling temperature. The battery applies unchanged to any dilemma set and binary format: measuring what a model values requires crossing the frames of the question, not asking once.
Read the full abstract

Computation and Language

Most LLM Conformity Needs No Speaker: Measuring the Speaker-Free Floor in Peer-Pressure Benchmarks

Yibo Hu and Jiaming Qu

LLM conformity is often used to describe cases where a model changes a correct answer toward a peer or group response. We show that most of this apparent conformity survives even after the peer is removed. The reason is a confound: standard conformity prompts mix two cues at once, the presence of a speaker and the repeated wrong answer itself. Existing benchmarks vary these cues together, so they cannot tell how much of the revision actually depends on the speaker. We introduce a no-source condition: the same asserted answer with the explicit speaker removed. Across six open-weight LLMs and seven QA and reasoning datasets, this condition alone causes harmful revision in $66.5\%$ of initially correct cases, compared with $10.3\%$ under a plain re-ask. The effect also remains when the repeated answer is paraphrased and when answer options are hidden in an open-ended setting. Source framing mainly modulates this floor: expert-panel framing raises it, while minimal person labels do not reliably raise it. When models flip, they are usually confidently wrong, and simple recalibration does not recover the original answer. Source attribution still matters, but it should be measured as an increment above this speaker-free floor. The methodological lesson is that conformity benchmarks should first measure what remains after the speaker is removed; without this step, benchmarks may mistake repeated text for social influence.
Read the full abstract

Computation and Language · Cross-listed from cs.MA

Decision Protocols in Multi-Agent Large Language Model Conversations

Lars Benedikt Kaesberg

Improving the task performance of Large Language Models (LLMs) is essential, yet scaling these models faces significant challenges such as diminishing returns and high costs. Multi-Agent Systems (MAS) offer a promising solution by distributing tasks among specialized agents to improve the overall task performance. This can reduce training costs at the expense of increased test time due to the discussion and decision-making process. The decision protocol is a critical component of MAS because it specifies how multiple agents collaborate to create a final solution. This thesis introduces the Multi-Agent LLM (MALLM) framework, which implements and evaluates various decision protocols, namely voting, consensus, and judge decision mechanisms, to simulate multi-agent discussions for conversational task solving. Unlike previous work that used a single decision protocol or tested them on limited datasets, this study systematically examines their impact on a diverse set of tasks, ranging from knowledge-based datasets (MMLU, MMLU-Pro, GPQA) and logic-based datasets (StrategyQA, MuSR, Math-lvl-5, SQuAD 2.0). The results indicate that consensus protocols excel in knowledge-intensive domains while voting and judge protocols are more effective for logic-based tasks. Increasing response diversity through independent solution generation improves decision quality, while changes in information access during the decision process have minimal impact.
Read the full abstract

Computation and Language · Cross-listed from cs.LG

Weak-to-Strong Generalization via Direct On-Policy Distillation

Shiyuan Feng, Huan-ang Gao, Haohan Chi et al. (+7 more)

Reinforcement learning with verifiable rewards (RLVR) is a powerful recipe for improving language-model reasoning, but it is expensive to repeat on every new strong model because the target model must generate many rollouts during training. As models scale, post-training itself becomes a bottleneck. We study a weak-to-strong alternative: run RL on a smaller model where rollouts are cheaper, then reuse what that RL run learned to improve a stronger target model. Directly distilling the post-RL weak teacher is not enough, because the teacher's final policy mixes useful RL gains with the limitations of the smaller model. We propose Direct On-Policy Distillation (Direct-OPD), which transfers the teacher's RL-induced policy shift instead. Direct-OPD compares the post-RL teacher with its own pre-RL reference and treats their log-ratio as a dense implicit reward for the student. In plain terms, the checkpoint pair tells us which actions RL made the weak model more or less likely to take, and Direct-OPD applies that signal on the stronger student's own on-policy states. This directly reuses the weak model's RL supervision signal without running sparse-reward RL on the target model. Empirically, Direct-OPD consistently leverages weaker teachers to improve stronger target models; notably, it boosts Qwen3-1.7B from 48.3% to 58.3% on AIME 2024 in just 4 hours on 8 A100 GPUs. It outperforms step-matched direct RL and enables the sequential composition of multiple policy shifts. Our results show that RL outcomes can be reused across model scales as implicit reward signals, not merely as final models to imitate.
Read the full abstract

All authors Shiyuan Feng, Huan-ang Gao, Haohan Chi, Hanlin Wu, Zhilong Zhang, Zheng Jiang, Bingxiang He, Wei-Ying Ma, Ya-Qin Zhang, Hao Zhou

Computation and Language · Cross-listed from cs.AI

LLM-as-a-Verifier: A General-Purpose Verification Framework

Jacky Kwok, Shulu Li, Pranav Atreya et al. (+6 more)

Scaling pre-training, post-training, and test-time compute have become the central paradigms for improving the capabilities of LLMs. In this work, we identify verification, the ability to determine the correctness of a solution, as a new scaling axis. To unlock this and demonstrate its effectiveness, we introduce LLM-as-a-Verifier, a general-purpose verification framework that provides fine-grained feedback for agentic tasks without requiring additional training. Unlike standard LM judges that prompt LLMs to produce discrete scores for candidate solutions, LLM-as-a-Verifier computes the expectation over the distribution of scoring token logits to generate continuous scores. This probabilistic formulation enables verification to scale along multiple dimensions: (1) score granularity, (2) repeated evaluation, and (3) criteria decomposition. In particular, we show that scaling the scoring granularity leads to better separation between positive and negative solutions, resulting in more calibrated comparisons. Moreover, scaling repeated evaluation and criteria decomposition consistently lead to additional gains in verification accuracy through variance and complexity reduction. We further introduce a cost-efficient ranking algorithm for selecting the best solution among candidates using the verifier's continuous scores. LLM-as-a-Verifier achieves state-of-the-art performance on Terminal-Bench V2 (86.5%), SWE-Bench Verified (78.2%), RoboRewardBench (87.4%), and MedAgentBench (73.3%). Beyond verification, the fine-grained signals from LLM-as-a-Verifier can also serve as a proxy for estimating task progress. We build an extension for Claude Code, enabling developers to monitor and improve their own agentic systems. Finally, we show that LLM-as-a-Verifier can provide dense feedback for RL, improving the sample efficiency of SAC and GRPO on robotics and mathematical reasoning benchmarks.
Read the full abstract

All authors Jacky Kwok, Shulu Li, Pranav Atreya, Yuejiang Liu, Yixing Jiang, Chelsea Finn, Marco Pavone, Ion Stoica, Azalia Mirhoseini

Computation and Language · Cross-listed from cs.LG

What Does a Discrete Diffusion Model Learn? One Reverse Process in Three Coordinates

Rodrigo Casado Noguerales, Bernhard Schölkopf, Thomas Hofmann et al. (+1 more)

What does a discrete diffusion model actually learn? The literature trains denoisers, bridge plug-ins, and concrete scores, each deriving its own ELBO objective under different conventions. This fragmentation makes the object being learned unclear and ELBO values difficult to interpret and compare. All three methods learn the same object in different coordinates: the exact time-reversal of the noising process. We first prove the Oracle Distance identity: for any noising process, the negative ELBO is not just a bound on the likelihood; it equals the data entropy plus the path KL divergence from the reversed noising process to the learned one. Next, for token-factorizing noise, we show that denoiser, bridge plug-in, and concrete score are three coordinates of this reverse process, derive exact conversions between them, and identify the probability distribution learned by each. In particular, contrary to its standard interpretation, the bridge plug-in learns not a denoiser but the cavity distribution: the clean token given the noisy sequence excluding its own position. We confirm the theory by training all three coordinates with masked, uniform, and GIDD kernels on OpenWebText under a shared recipe, and by training uniform-diffusion solvers on four logic tasks. The learned coordinates agree after conversion, while confusing the cavity with the denoiser can severely degrade sampling quality. Together with exact boundary-term formulas, closed-form NELBO calibrations at initialization, and a clock interpretation of time schedules, these results provide a conceptual framework and a mechanical recipe for deriving, implementing, and comparing discrete diffusion processes.
Read the full abstract

All authors Rodrigo Casado Noguerales, Bernhard Schölkopf, Thomas Hofmann, Aran Raoufi

Computation and Language · Cross-listed from cs.RO

GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks

Kaiyuan Chen, Shuangyu Xie, Letian Fu et al. (+21 more)

For robots to work reliably in commercial and industrial applications, recent advances in agentic coding systems have potential to combine interpretable robot programming with the open-world adaptability of model-free policies. We focus on "Variational Automation" (VA), a class of tasks that have larger variations in object geometry and pose than fixed automation. Model-free policies often struggle to close the reliability gap for VA tasks, which must be executed persistently and reliably in commercial and industrial applications. Motivated by prior work on Task and Motion Planning (TAMP) and the Robot Operating System (ROS), we introduce Graph-as-Policy (GaP), a multi-agent coding harness that generates directed computation graphs with perception, planning, and control nodes from a Modular Open Robot Skill Library (MORSL). GaP generates an internal simulation environment to rehearse task instances with different graphs in parallel to iteratively refine the graph structure and parameters to improve success rates and throughput. Evaluation with 8 new open VA task benchmarks, 4 in simulation and 4 in real-world, suggests that GaP can achieve success rates that significantly outperform baselines. Details, code, and data: https://graph-robots.github.io/gap
Read the full abstract

All authors Kaiyuan Chen, Shuangyu Xie, Letian Fu, Justin Yu, William Pacini, Sandeep Bajamahal, Hudson Kim, Jaimyn Drake, Daehwa Kim, Haoru Xue, Jonathan Francis, Christian Juette, Peter Schaldenbrand, Muhammet Yunus Seker, Ruwan Wickramarachchi, Uksang Yoo, Guanzhi Wang, Adithyavairavan Murali, Balakumar Sundaralingam, S. Shankar Sastry, Yuke Zhu, Linxi "Jim" Fan, Spencer Huang, Ken Goldberg

Computation and Language

SPEARBench: A Benchmark for Naturalness Evaluation in Streaming Speech-to-Speech Language Models

Thomas Thebaud, Yuzhe Wang, Hao Zhang et al. (+5 more)

Streaming speech-to-speech language models aim to answer spoken queries directly with synthetic speech. However, standard speech and text benchmarks do not capture whether these systems behave naturally in conversations, where timing, turn-taking, prosody, interpersonal stance, language and dialect consistency, and relationship-aware appropriateness jointly shape perceived quality. We introduce SPEARBench, a benchmark for evaluating naturalness in speech-to-speech language models from question-answer interactions. SPEARBench constructs controlled dialogue prompts from the Seamless Interaction corpus, runs inference across multiple models, and evaluates generated answers using a multidimensional protocol that covers response latency, interruptions, speech quality, ASR robustness, language and dialect consistency, emotional naturalness, interpersonal stance, and explainable distributional baselines. The benchmark includes original human answers as a reference condition and reports results for several contemporary models. Results show that current models can achieve high signal-level quality and low ASR error while still differing from human conversational behavior in latency, overlap, dialect preservation, emotional adaptation, and interpersonal stance dynamics.
Read the full abstract

All authors Thomas Thebaud, Yuzhe Wang, Hao Zhang, Sathvik Manikantan Napa Ugandhar, Ashish Hallur, Georgi Tinchev, Venkatesh Ravichandran, Laureano Moro-Velazquez

Computation and Language

REDDIT: Forgetting-Resistant Correction of Timestamp Drift in ASR via Replay-Based Distribution Editing

Cheng-Kang Chou, Ming-To Chuang, Ke-Han Lu et al. (+2 more)

Modern autoregressive ASR systems can emit timestamps as decoded tokens, enabling timestamped transcription without frame-level aligners or inference-time post-processing. We show that these generated timestamps can drift across long non-speech spans: the transcript may remain plausible, but the decoded time axis drifts away from the audio. We study this non-speech-induced timestamp drift with self-built gap and long-gap benchmarks across 15 evaluated timestamp-producing ASR and audio-language systems. Naive timestamp-corrected fine-tuning improves alignment but can severely degrade non-target ASR behavior, exposing a forgetting problem. We propose REDDIT(REplay-based Distribution eDITing), a lightweight two-stage post-training framework that corrects timestamps while avoiding this catastrophic forgetting: it first edits timestamp targets under the model's own replayed decoder context while matching the frozen base distribution on non-timestamp tokens, then applies a short edited-prefix refinement stage. In this framework, we construct correction supervision without human transcripts or human timestamp annotations by combining VAD-trimmed speech spans with inserted non-speech gaps and known concatenation offsets. On Whisper-tiny, 34.9 hours of targeted correction audio used and only 1.6% of model parameters updated, raising long-gap mIoU from 38.7% to 95.0% and reducing mixed-gap out-of-domain AAS from 2752 ms to 223 ms while preserving CV-en MER at 41.3% (versus 524.2% for ordinary SFT decoder tuning).
Read the full abstract

All authors Cheng-Kang Chou, Ming-To Chuang, Ke-Han Lu, Chan-Jan Hsu, Hung-yi Lee

Computation and Language

Faithfulness to Refusal: A Causal Audit of Neuron Selectors

Ananth Eswar, Pratinav Seth, Utsav Avaiya et al. (+1 more)

Attribution scores increasingly identify which neuron rows of a language model matter for applications such as pruning, interpretability, and editing for safety, yet whether they identify causally important rows is rarely tested directly. We address this with two paired audits built on one-shot neuron-row zeroing. We first audit selectors at the language-modeling level: attribution methods substantially outperform activation and magnitude-based baselines at identifying dispensable rows across five LLMs. We then adapt the same intervention into a behavior test by driving it with a contrastive harmful-versus-benign signal; the attributed rows are sufficient to install refusal on hate and crime while keeping benign over-refusal low and preserving language model fluency, and specific in that layer-matched random controls at the same depths fail. Highly rank-stable selectors can be among the least causally valid. Refusal moreover lives in a redundant subspace, where different attribution methods install it through largely disjoint row sets, so the recovered edit is one realization of a sufficient set rather than a unique mechanism. Together, these findings show that rank-stability proxies miss the kinds of selector failures a direct causal audit can surface.%
Read the full abstract

All authors Ananth Eswar, Pratinav Seth, Utsav Avaiya, Vinay Kumar Sankarapu

Computation and Language · Cross-listed from cs.CR

Selective Disclosure Watermarking for Large Language Models

Xuyang Chen, Xiang Li, Yangxinyu Xie et al. (+1 more)

Watermarking methods embed imperceptible and verifiable signals into text generated by large language models (LLMs). Existing approaches include zero-bit schemes for distinguishing synthetic text from human writing and multi-bit schemes for embedding metadata. However, current multi-bit watermarking methods do not allow selective disclosure: verifying any part of the watermark requires revealing the entire embedded message. This lack of control leads to unnecessary information exposure and raises privacy concerns. We propose Hierarchical Vocabulary Routing (HeRo), a watermarking framework that enables selective disclosure of embedded metadata. The method recursively partitions the vocabulary and distributes watermark information across hierarchical layers, so that different verifiers can decode only the portions of the payload corresponding to their access level. We show that the proposed scheme preserves the unbiasedness of the underlying sampling process and thus maintains text quality. Experiments demonstrate that our framework supports fine-grained access control while achieving high detection accuracy and low latency. Code is available at https://github.com/xuyangc03/hero-watermark.
Read the full abstract

All authors Xuyang Chen, Xiang Li, Yangxinyu Xie, Qi Long

Computation and Language

How Much is Left? LLMs Linearly Encode Their Remaining Output Length

Mohamed Amine Merzouk, Dmitri Carpov, Mirko Bronzi et al. (+2 more)

Large language models generate one token at a time, yet their responses show remarkably consistent length structure: step-by-step solutions converge in predictable token counts, retrievals stop after a few sentences, retractions extend responses by measurable amounts. We ask whether the model carries an internal estimate of how much response remains. Training minimal-capacity linear probes on frozen hidden states of three open-weight 7-8B models across seven completion-style datasets, we find three converging pieces of evidence. First, total response length is linearly decodable from the prompt's last hidden state alone, before any output is emitted. Second, probe directions trained on natural-language datasets transfer broadly, including to controlled synthetic completions never seen in training, outperforming a statistical baseline; the converse direction generally fails, and this asymmetry is itself informative. Third, on curated high-loss completions, the probe's per-position estimate shifts upward at the moment the model retracts and restarts a partial solution, a directional behavior no position-only predictor can reproduce (qualitative, not aggregate). We frame this as approximate estimation of remaining generation length, distinct from exact-counting impossibility results for transformers, and interpret it as evidence that LLMs maintain a plan-like internal representation of output length (decodable, not necessarily used causally).
Read the full abstract

All authors Mohamed Amine Merzouk, Dmitri Carpov, Mirko Bronzi, Damiano Fornasiere, Adam Oberman

Computation and Language

SalAngaBhava: A Sinhala Market Dataset for Aspect-based Sentiment Analysis

Lakshani Galwatta, Nisansa de Silva, Sarangi Aththanayake et al. (+1 more)

Sentiment analysis has been a primary domain under Natural Language Processing (NLP) from its inception as it plays a vital role in both real-world and research applications. In high-resource languages, this has been extended a step further, and instead of predicting sentiment at the sentence level, models have been developed to detect more fine-grained sentiments at aspect level. However, in order to conduct this fine-grained Aspect-based Sentiment Analysis (ABSA), datasets annotated with aspects and sentiments toward the said aspects is required. Such datasets are lacking for low-resources languages among which, we can count Sinhala, an Indo-Aryan languages used primarily in Sri Lanka. In this work, we introduce, SalAngaBhava, a new Sinhala Aspect-based Sentiment Analysis dataset which contains Sinhala product reviews that are manually labeled with aspect terms and the associated sentiments (positive, negative, neutral). The data was collected from domain-relevant sources such as user-generated reviews and comments, and was annotated following carefully defined guidelines to ensure consistency and quality. The dataset consists of sentences and aspect-sentiment pairs, encompassing a considerable range of aspects from several domains. The analysis confirms that the dataset is well-structured and sufficiently balanced for ABSA research. This dataset can be used as a benchmark and facilitates further studies related to Sinhala natural language processing, and low-resource sentiment analysis tasks.
Read the full abstract

All authors Lakshani Galwatta, Nisansa de Silva, Sarangi Aththanayake, Adithya Galwatta

Computation and Language

Streaming Neural Speech Codecs through Time-Invariant Representations

Kélian Estève, Salima Mhdaffar, Mickael Rouvier et al. (+2 more)

Neural speech codecs are increasingly used as intermediate representations in codec-based speech generation systems. TiCodec introduces a factorized representation that separates time-varying speech content from time-invariant information through a Time-Invariant Representation Extraction (TIRE) module, potentially reducing the amount of information that must be modeled at the frame-level. In this work, we investigate the nature of the information captured by TIRE representations and their suitability for low-latency speech processing. Using a series of probing tasks, we analyze the influence of the encoder layer and show that intermediate layers capture complementary speaker- and environment-related information while containing little linguistic content. We further study several segment selection strategies for TIRE training and demonstrate that cross-file sampling improves the robustness of invariant representations. Based on these findings, we propose Dual-TIRE, a multi-level architecture that exploits the complementarity of different encoder layers and improves speech reconstruction quality and speaker similarity. Finally, we evaluate TiCodec in a streaming inference setting using successive 660ms processing blocks. Results show that streaming operation can be achieved without significant degradation in reconstruction performance, highlighting the potential of factorized neural codec representations for future low-latency speech generation systems.
Read the full abstract

All authors Kélian Estève, Salima Mhdaffar, Mickael Rouvier, Richard Dufour, Yannick Estève

Computation and Language

Progressive Refinement: An Iterative Pseudo-Labeling Approach for Mandarin-English Code-Switching ASR

Qu Yang, Cakra Wardhana and Tim Ng

Code-switching (CS), alternating languages within the same utterance, poses significant challenges for automatic speech recognition (ASR) due to limited CS training data. This paper applies an iterative pseudo-labeling training approach to CS-ASR for the first time, demonstrating its effectiveness in leveraging unlabeled data to improve CS-ASR performance. The approach comprises three phases: pseudo-label generation, two-stage bilingual model training, and iterative improvements. It begins by generating pseudo-labels from a large unlabeled corpus, creating a semi-supervised dataset. This dataset supports a two-stage training framework where the model is pre-trained and then fine-tuned on supervised CS data. Iterative refinements further enhance the model's accuracy in handling complex CS scenarios. Our approach significantly advances CS-ASR systems, achieving notable Mix Error Rate (MER) reductions on SEAME's devman (6.35%) and devsge (8.29%) subsets.
Read the full abstract

Computation and Language · Cross-listed from cs.CY

Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off

Hafsteinn Einarsson, Hafsteinn Birgir Einarsson, Jón Gunnar Ólafsson et al. (+1 more)

Public institutions increasingly use large language models (LLMs) to answer citizens' questions, often pairing a curated knowledge base with live web search, yet whether the sources behind these answers can be trusted has received little empirical scrutiny. We report a pre-launch expert evaluation of Evrópuvefur, an independent, government-funded service run by the University of Iceland that answers questions about the European Union, conducted as Iceland prepared for its referendum of 29 August 2026 on whether to resume EU accession talks. Five domain experts produced 551 evaluations of 449 AI-generated answers, scoring each against a seven-criterion quality rubric and, separately, flagging individual cited sources. We compared two retrieval paths: a curated local corpus (RAG) and open web search. In more than a third of the reviewed web-search answers (35%, 65 of 187), at least one cited source was flagged, almost always as untrustworthy or irrelevant; curated sources were flagged far less often and only for being out of date. Web search answered more questions, but at the cost of source quality; the curated corpus was trustworthy yet limited in coverage, and the model declined to respond when it fell short. The citation mix also passed over strong sources: across all 287 web-search answers, the system never cited RÚV, the public broadcaster and the country's most widely used news source. A companion prompt ablation shows how weak prompt-level steering is: a trusted-domain list in the system prompt raised the share of citations to listed domains only from 12% to 21%. Fluency and topical fit did not predict source trustworthiness. We argue that source trustworthiness is a measurable yet largely invisible dimension of information quality in public AI services, and we discuss transparency-oriented responses and their trade-offs.
Read the full abstract

All authors Hafsteinn Einarsson, Hafsteinn Birgir Einarsson, Jón Gunnar Ólafsson, Jón Gunnar {\TH}orsteinsson

Computation and Language · Cross-listed from cs.LG

Noisy-Channel Minimum Bayes Risk Decoding

Yusuke Sakai, Hidetaka Kamigaito and Taro Watanabe

Minimum Bayes Risk (MBR) decoding yields more robust and higher-quality text generation than maximum a posteriori (MAP) decoding by selecting hypotheses that maximize expected utility over sampled pseudo-references. However, there exists a discrepancy in the design: hypothesis selection calculates expected utility scores conditioned on given pseudo-references, while commonly used evaluation metrics, e.g., BLEU and COMET, are asymmetric. Therefore, it is important to consider both hypothesis-to-reference and reference-to-hypothesis directional effects. In this study, we introduce a noisy channel decomposition of MBR decoding that naturally incorporates bidirectional effects to account for these asymmetries. We decompose MBR decoding into four interacting components: hypothesis-to-reference likelihood, reference-to-hypothesis likelihood, hypothesis prior, and reference prior. This decomposition provides a unified interpretation of existing MBR variants and enables metric- and task-specific interpretability by isolating the contribution of each channel. Our comprehensive analysis reveals that channel-wise contributions exhibit distinct characteristics across metrics while remaining consistent across tasks, and suggests that appropriate channel weighting may lead to improvements over original MBR decoding.
Read the full abstract

Computation and Language

Unified Audio Intelligence Without Regressing on Text Intelligence

Zhifeng Kong, Sang-gil Lee, Jaehyeon Kim et al. (+17 more)

Audio intelligence involves understanding, reasoning about, and generating both audio and speech. In this work, we introduce Nemotron-Labs-Audex-30B-A3B (Audex), a unified audio-text LLM built on Nemotron-Cascade-2-30B-A3B, a strong text-only MoE LLM. Audex adopts a simple unified design with a single Transformer decoder: audio inputs are encoded and projected into the text embedding space, while text tokens and quantized audio output tokens are treated uniformly during generation. This architecture enables strong audio-text fusion, seamless multimodal generation, and compatibility with standard LLM training and inference infrastructure. For training, we meticulously curate audio-text datasets comprising 157.4B audio tokens and 320.5B text tokens. We apply multi-stage supervised training on these datasets, followed by text-only Cascade RL and multi-domain on-policy distillation. Audex delivers state-of-the-art audio understanding, speech recognition and translation, text-to-speech, audio generation, and speech-to-speech generation, while preserving very compelling reasoning, alignment, knowledge, long-context, and agentic capabilities of its text-only LLM backbone with marginal or no regression. We release the model checkpoints to facilitate open research.
Read the full abstract

All authors Zhifeng Kong, Sang-gil Lee, Jaehyeon Kim, Boxin Wang, Zihan Liu, Sungwon Kim, Yang Chen, Arushi Goel, Rajarshi Roy, Wenliang Dai, Zhuolin Yang, Yangyi Chen, Dongfu Jiang, Sreyan Ghosh, Tuomas Rintamaki, Andrew Tao, Jonathan Raiman, Mohammad Shoeybi, Bryan Catanzaro, Wei Ping

Computation and Language

RABBiT: Rapidly adaptive BOLD foundation model via brain-tuning for accurate zero-shot and few-shot prediction of speech-elicited responses in the brain

Omer Moussa and Mariya Toneva

Language understanding in the brain is context-dependent, varying across experimental stimuli and individuals, which makes it difficult to build computational models that generalize across both. This calls for a foundation model of language-evoked brain activity that can capture shared structure while adapting efficiently to new participants and inputs. We introduce RABBiT (Rapidly Adaptive BOLD foundation model via BraIn-Tuning), a compact audio-to-fMRI encoder designed for accurate zero- and few-shot prediction. A comprehensive evaluation on 324 participants across multiple unseen fMRI datasets shows that RABBiT enables accurate zero-shot prediction of fMRI responses to natural speech across auditory and language-selective regions, surpassing the SOTA foundation model for fMRI and predictions based on group averages. With as little as 10 minutes of participant-specific data, RABBiT further improves performance via parameter-efficient tuning, substantially outperforming per-participant linear models. RABBiT's performance is driven by two key innovations: (1) learned region-specific attention, and (2) a decomposition of brain responses into shared and subject-specific components, combined with a brain-tuned speech backbone. In addition to supporting strong predictive accuracy, the structured, region-specific representations that RABBiT learns enable interpretability. By eliminating the need for extensive per-participant data and model fitting, RABBiT enables scalable population-level analyses of language in the human brain. We make the code available at https://github.com/bridge-ai-neuro/rabbit.
Read the full abstract

Computation and Language

EdgeBench: Unveiling Scaling Laws of Learning from Real-World Environments

Deyao Zhu, Xin Zhou, Shengling Qin et al. (+44 more)

Pretraining scaling laws reveal that model capability improves predictably with data and compute. But learning from real world environments after deployment remains far less understood. Analyzing roughly 38,000 hours of agent interaction with the environment across 134 real world tasks, we find, to the best of our knowledge, the first evidence that overall performance during environment learning follows a log-sigmoid scaling law with remarkably high precision, reaching R^2 = 0.998. Across model generations, we also find that agent learning speed roughly doubles every three months. This discovery stems from EdgeBench, a suite of 134 real world tasks with ultra-long horizons, spanning scientific discovery, software engineering, combinatorial optimization, professional knowledge work, formal mathematics, and interactive games. Each task sustains at least 12 hours of continuous agent operation under rich, multilevel feedback, and is built through substantial expert effort. We publicly release 51 tasks and our full evaluation framework to accelerate the study of how agents learn from real world experience.
Read the full abstract

All authors Deyao Zhu, Xin Zhou, Shengling Qin, Xuekai Zhu, Hangliang Ding, Shu Zhong, Zixin Wen, Zhonglin Xie, Chenhui Gou, Linxuan Ren, Yueyang Wang, Junfeng Zhong, Rui Liu, Tian Gao, Yangguang Lin, Jingyuan Zhang, Maojia Song, Xuan Qi, Jinhong Wu, Chenyang Zhang, Yinzhu Piao, Ziru Niu, Hongbin Lin, Lingxiang Meng, Peng Tang, Chengyao Tang, Shanyu Wu, Huanyu Zheng, Yu Liu, Liya Zhu, He Wang, Ming Ding, Ziyu Wan, Hao Liu, Sibo Wang, Haotian Zhu, Xintian Zhang, Nan Chai, Yipeng Liu, Panhao Lai, Sihang Yuan, Zixin Su, Ge Zhang, Wangchunshu Zhou, Yantao Du, Wenhao Huang, Guang Shi

Computation and Language · Cross-listed from cs.AI

DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation

Xin Cheng, Xingkai Yu, Chenze Shao et al. (+30 more)

Speculative decoding accelerates Large Language Model (LLM) inference by decoupling draft generation from target verification. While recent parallel drafters efficiently propose long token sequences in a single forward pass, they suffer from rapid acceptance decay due to a lack of inter-token dependencies. Furthermore, indiscriminately verifying these extended blocks wastes critical batch capacity on tokens with high rejection risks, severely degrading throughput in high-concurrency serving systems. We introduce DSpark, a speculative decoding framework that unifies high-throughput parallel generation with adaptive, load-aware verification. To maintain draft quality, DSpark utilizes a semi-autoregressive architecture, coupling a parallel backbone with a lightweight sequential module, to introduce intra-block dependency modeling and mitigate suffix decay. To optimize system efficiency, DSpark employs confidence-scheduled verification, dynamically tailoring the verification length for each request based on estimated prefix survival probabilities and engine-specific throughput profiles. On offline benchmarks across diverse domains, DSpark substantially improves the accepted length over state-of-the-art autoregressive and parallel drafters. When deployed within the DeepSeek-V4 serving system under live user traffic, DSpark successfully mitigates verification waste. Compared to the established production baseline (MTP-1), DSpark accelerates per-user generation speeds by 60 to 85 percent at matched throughput levels. More importantly, by preventing severe throughput degradation under strict interactivity constraints, it enables performance tiers that were previously unattainable, shifting the Pareto frontier of our serving system.
Read the full abstract

All authors Xin Cheng, Xingkai Yu, Chenze Shao, Jiashi Li, Yunfan Xiong, Yi Qian, Jiaqi Zhu, Shirong Ma, Xiaokang Zhang, Jiasheng Ye, Qinyu Chen, Chengqi Deng, Jiping Yu, Damai Dai, Zhengyan Zhang, Yixuan Wei, Yixuan Tan, Wenkai Yang, Runxin Xu, Yu Wu, Zhean Xu, Xuanyu Wang, Muyang Chen, Rui Tian, Xiao Bi, Zhewen Hao, Shaoyuan Chen, Huanqi Cao, Wentao Zhang, Anyi Xu, Huishuai Zhang, Dongyan Zhao, Wenfeng Liang

Computation and Language · Cross-listed from cs.CY

When Agents Lie: Premeditation, Persistence, and Exploitation in Repeated Games

Jerick Shi, Terry Jingcheng Zhang, Bernhard Schölkopf et al. (+2 more)

As large language models are deployed as autonomous agents that communicate intentions before acting, a critical safety question is whether agents that publicly commit to actions will honor those commitments. We place LLM agents in repeated $n$-player games with a three-stage protocol that separates private intent, public announcement, and final action, allowing us to identify whether each deviation from a stated announcement was already planned during private deliberation. Evaluating three frontier models across six games in homogeneous and heterogeneous groups over 10 rounds, we report two findings. First, when agents deviate from their announcements, the deviation is predominantly already stated in their private plan (exceeding 90% in the highest-deception conditions), yet this is not a fixed model property: the same model ranges from perfect honesty to near-total deviation across games. Second, different models interpret announcements incompatibly, some as binding commitments and others as cheap talk, producing payoff gaps that emerge in Round~0 and persist across all 10 rounds. Systems that combine models from different providers therefore cannot assume shared announcement semantics and require empirical testing of model interactions before deployment.
Read the full abstract

All authors Jerick Shi, Terry Jingcheng Zhang, Bernhard Schölkopf, Vincent Conitzer, Zhijing Jin

Computation and Language · Cross-listed from cs.LG

Localized LoRA-MoE: Block-wise Low-Rank Experts With Adaptive Routing

Babak Barazandeh, Subhabrata Majumdar, Vinay Prithyani et al. (+1 more)

Large Language Models (LLMs) and high-dimensional perception networks increasingly rely on parameter-efficient fine-tuning (PEFT) to adapt to diverse operational contexts. However, standard methods like LoRA are structurally limited by a monolithic bottleneck, making them highly susceptible to gradient warfare. Interleaved multi-task streams may trigger destructive optimization feedback, collapsing adapter weights into unspecialized averages. While recent spatial partitioning methods have introduced block-wise isolation, they remain trapped in static topologies, unable to adapt to dynamic task-switching or environmental sensor failure. In this work, we introduce Localized LoRA-MoE, a unified framework that fuses localized spatial blocking with dynamic, context-conditioned routing. We propose and evaluate two novel architectural paradigms: Block-Wise LoRA-MoE (Centralized Macro-Routing), which modulates the entire structural grid via a monolithic context signal, and Cell-Wise LoRA-MoE (Decentralized Micro-Routing), which empowers every coordinate cell in the matrix grid with autonomous, localized expert gating. Through a comprehensive suite of benchmarks, ranging from high-dimensional SVD matrix simulations and real-world tabular transformations to spatial vision perception under sensor degradation, we demonstrate that both architectures resolve optimization deadlocks inherent in static baselines. Our empirical results establish that decentralized cell-level gating achieves complete statistical parity with an omniscient global coordinator, providing a robust "gradient firewall" that protects surviving pathways from fault-propagated corruption. Our proposals consistently outperform static baselines, offering a scalable and parameter-efficient solution for dynamic model adaptation across granular coordinate fields and shifting operational regimes.
Read the full abstract

All authors Babak Barazandeh, Subhabrata Majumdar, Vinay Prithyani, George Michailidis

Computation and Language

Rating the Pitch, Not the Product: User Evaluations of LLMs Reflect Expectations More Than Performance

Robert Morabito, Tyler McDonald, Charitra Viswanath et al. (+4 more)

Imagine two users interact with the same LLM. One has been told it is the cutting-edge flagship model; the other, an older, weaker model. They walk away with markedly different ratings of its usefulness and intelligence, yet they used the same model. In a controlled study, 162 participants each used one of six LLMs from two families across three collaborative tasks, after first viewing a landing page that matched, overstated, or understated their model's true capability. This pre-interaction framing shifted user opinions and interaction behavior while task performance did not. Oversold users rated the model more favorably and used more directive prompting, while Undersold users wrote longer, more collaborative prompts. The quality of what users and the model produced together depended only on the model's true capability, not on what users were told. Participants' change in model impressions after use, measured across two impression measures, was not predicted by task performance ($\beta = -0.01$ and $0.11$, both n.s.), but by whether the model met users' expectations ($\beta = 0.47$ and $0.50$, both $p < .001$) and how confident they felt working with it ($\beta = 0.47$ and $0.36$, both $p < .001$). After interaction, users are still rating the pitch, not the product: user-elicited LLM evaluations, including the preference data driving public leaderboards, measure expectation management at least as much as the model itself.
Read the full abstract

All authors Robert Morabito, Tyler McDonald, Charitra Viswanath, Angel Hsing-Chi Hwang, Susanne Gaube, Jad Kabbara, Ali Emami

Computation and Language

MIRAGE: Defending Long-Form RAG Against Misinformation Pollution

Saadeldine Eletter, Ruihong Zeng, Yuxia Wang et al. (+3 more)

Retrieval-Augmented Generation (RAG) improves factuality by grounding LLMs in external evidence, but real-world retrieval is often polluted: semantically relevant passages may contain subtle misinformation, misleading framings, or fabrications. We introduce MIRAGE, a training-free, model-agnostic defense for long-form RAG. MIRAGE builds an NLI-based cross-document claim graph and applies a Defended-Claims Gate to either condition generation on a consistent, multi-source supported subset or to block retrieval and answer parametrically. We also release a minimal-edit pollution protocol spanning four perturbation families (Unambiguous, Conflicting, Misleading, Fabricated) to construct matched clean, mixed, and fully polluted evaluation regimes. Across four long-form QA benchmarks and multiple commercial and open-weight LLMs, pollution severely degrades vanilla RAG, while MIRAGE consistently restores factuality under mixed and fully polluted evidence and outperforms prior robust-RAG methods. Our implementation and datasets are available at https://github.com/SaadElDine/MIRAGE.
Read the full abstract

All authors Saadeldine Eletter, Ruihong Zeng, Yuxia Wang, Maxim Panov, Aleksandr Rubashevskii, Preslav Nakov

Computation and Language

Beyond Independent Labels: Schwartz-Geometry Decoding for Human Value Detection

Víctor Yeste and Paolo Rosso

Human value detection is commonly formulated as sentence-level multi-label classification over the 19 refined Schwartz values, typically predicted as independent labels. Schwartz theory, however, describes them as a circular motivational continuum, in which adjacent values are compatible and opposing values are in tension. We ask whether this structure can be operationalized as an explicit output-space geometry and used as a soft bias rather than a hard constraint. On a DeBERTa-v3-base classifier, we compare two ways of injecting it: training-time geometry-aware objectives and a post-hoc Schwartz-aware energy decoder that scores whole label sets jointly. Across five seeds, training-time geometry gives only limited gains-no larger for the true continuum than for a random ordering-whereas the decoder makes label sets more coherent with the continuum-on theory-aware coherence metrics we introduce-at no cost to Macro-F1 or Micro-F1 (held fixed by its selection rule). The gain is specific to the true Schwartz ordering: it does not appear for a random permutation or an empirical co-occurrence graph through the identical decoder. A bounded Qwen2.5-72B-Instruct diagnostic shows that supplying the continuum at inference shifts behavior but does not match supervised structured prediction. Theory-aware decoding thus offers a lightweight, controllable way to make value detection faithful to its label space.
Read the full abstract

Computation and Language

Multi-Large Language Model Orchestrated Severity Assessment of Clinical Records (MOSAIC)

Manuela Del Castillo Suero, Arnault-Quentin Vermillet, Nicole Sonne Heckmann et al. (+2 more)

Background: Disease severity is a multidimensional construct difficult to capture with rule-based approaches in Electronic Healthcare Records (EHR). Agentic large language model (LLM) systems could synthesise clinical evidence and reason over EHRs, but remain unevaluated for this task. Methods: MOSAIC is a two-phase agentic LLM framework for severity phenotyping, using type 2 diabetes (T2D) as a proof-of-concept. MOSAIC was evaluated on a synthetic cohort (SyntheticMass; open-weight N = 4,886; closed-weight N = 200) against three algorithmic ground truths (DCSI, DiSSCo, Cooper) and against all-cause mortality and incident complications. Open-weight (locally deployable) and proprietary pipelines were also compared. Results: The generated framework spanned domains absent from the comparators, including biomarker-based glycaemic staging, beta-cell function, and social determinants of health. Open-weight MOSAIC matched the proprietary pipeline (closed- vs open-weight weighted kappa = 0.773) and reached moderate agreement with Cooper (kappa = 0.597) and DCSI (kappa = 0.534) and fair agreement with DiSSCo (kappa = 0.320). Agent-based (Type 1) tiers showed significant separation of all-cause mortality (log-rank p < 0.001; crude hazard ratios 1.6-2.4 for non-Baseline tiers), with non-monotonic separation at the upper tiers, and an inverse gradient for incident complications (log-rank p < 0.001) consistent with depletion of susceptibles. Agentic classification also diverged from deterministic execution of the same rubric (MOSAIC Frozen; kappa = 0.428), indicating reasoning beyond fixed rules. Conclusion: MOSAIC shows agentic LLM systems can generate and apply clinically meaningful severity phenotypes from structured EHR data in T2D. Extending it to other diseases with similarly multidimensional severity warrants further research.
Read the full abstract

All authors Manuela Del Castillo Suero, Arnault-Quentin Vermillet, Nicole Sonne Heckmann, Darmendra Ramcharran, Maurizio Sessa

Computation and Language

Knowledge Knows, Verbalization Tells: Disentangling Latent Directions for Mathematical Solvability in LLMs

Nikolaos Xiros, Maria-Eleni Zoumpoulidi and Georgios Paraskevopoulos

Although LLMs have made significant progress in mathematical reasoning, determining whether a mathematical problem is solvable remains a fundamental yet challenging capability. While recent studies have probed internal representations of model solvability beliefs, verbalization has primarily been studied behaviorally rather than as an internal representation, limiting its analysis and manipulation. We address this gap by separately probing representations of solvability knowledge and verbalization, allowing us to disentangle the two within model hidden states. Across multiple LLMs, we show that knowledge and verbalization are encoded as distinct, linearly decodable representations and that fabrication is primarily associated with changes in verbalization rather than the underlying knowledge. Prompting with unsolvability cues reduces fabrication primarily by shifting verbalization, while activation steering demonstrates that these representations can be echanistically manipulated to improve model abstention.
Read the full abstract

Computation and Language

The syntax of wh-agreement in Yemeni Ibbi Arabic

Ashraf Naji and Mohammed Q. Shormani

This article tackles an important phenomenon in the syntax of Yemeni Ibbi Arabic (YIA), viz., wh-agreement, a phenomenon common to several languages including Greek, Indonesian, Lubukusu, Irish, etc. In YIA, wh-agreement manifests itself via agreement inflections on the Wh-Op, C, T/V, v. To account for this phenomenon, we propose an Agree across phases (AAP) approach anchored in the mechanism of Feature Inheritance (FI) in which Agree as MATCHING (AM) is a bit separated from feature valuation (FV). AM concerns Cs/vs, but FV Ts/Vs. Analyzing the agreement patterns observed between Wh-Op(erators), functional heads (precisely C, (T), v), and verbal complexes, we argue that the suffixes -eh, -uh, -nen, -um, having undergone grammaticalization process from Stannard Arabic (SA) third person pronouns, function as morphological marking of wh-agreement. Findings indicate that YIA data offer a unique empirical contribution to generative syntax, specifically concerning wh-agreement in this dialect operating via MATCHING mechanism. Our proposal straightforwardly accounts for wh-agreement cross-linguistically. This study provides further evidence that incorporating under-investigated typology provides further support for the universality of Universal Grammar (UG) by revealing how specific I-language operations reflect deeper, invariant principles of human language architecture. It concludes that the wh-agreement mechanism in YIA is more morphosyntactically robust than in languages such as Greek, Indonesian, Palauan, and Irish, providing compelling evidence for AAP as a UG approach to long-distance dependencies.
Read the full abstract

Computation and Language · Cross-listed from cs.LG

Train Smarter, Not Longer: Memorization-Guided Data Reuse for Efficient LLM Training

Jingwei Zuo, Cong Zeng, Ilyas Chahed et al. (+6 more)

The training paradigm of large language models has shifted from traditional one-pass training to multi-epoch training, as reasonable reuse of limited high-quality data can improve both model performance and sample efficiency. Meanwhile, excessive repetition introduces the risk of overfitting and diminishing returns. Determining when and how to reuse data effectively thus emerges as a natural but under-explored question. Through a novel observation of model's "Memorization Window" signals derived from loss retention dynamics and downstream evaluation scores, we propose "Memorization-guided Data Reuse", a training paradigm that adaptively determines when and how data should be reused, enabling principled decisions on the number of training epochs and the scheduling of data replays. Our preliminary experiments reveal a consistent memorization-driven regime: performance continues to improve with repetition far beyond current practice (e.g., the commonly cited four-epoch limit). While a full scheduler remains future work, these insights provide a foundation for memorization-aware training schedules, helping to determine reuse budgets and move toward training LLMs smarter rather than longer with limited high-quality data.
Read the full abstract

All authors Jingwei Zuo, Cong Zeng, Ilyas Chahed, Maksim Velikanov, Dhia Eddine Rhaiem, Pasquale Balsebre, Abhay Kumar, Younes Belkada, Hakim Hacid