Computation and Language · Cross-listed from cs.AI
Beyond Symbolic Solving: Multi Chain-of-Thought Voting for Geometric Reasoning in Large Language Models
Md. Abu Bakor Siddique, Shahrin Hossain, Sadman Ahmed Siam et al. (+3 more)
Geometric Problem Solving (GPS) remains at the heart of enhancing mathematical
reasoning in large
language models because it requires the combination of diagrammatic understanding, symbolic manipulation and logical inference. In existing literature, researchers have chiefly focused on synchronising the diagram descriptions with text literals and solving the problem. In this vein, they have either taken a neural, symbolic or neuro-symbolic approach. But this solves only the first two of the requirements, namely diagrammatic understanding and symbolic manipulation, while leaving logical inference underdeveloped. The logical inference is often limited to one chain-of-thought (CoT). To address this weakness in hitherto existing models, this paper proposes MARS-GPS, that generates multiple parallel
reasoning rollouts augmented with Python code execution for numerical verification, ranks them using
token-level
entropy as a confidence signal, and aggregates answers through a multi-stage voting and self-verification pipeline. Empirical results show that MARS-GPS with 8 parallel rollouts achieves 88.8% on Geometry3K, a nearly +11% improvement over the prior state-of-the-art, with accuracy scaling consistently as the number of rollouts increases from 1 to 16 (+6.0% on ablation subset). We provide our code and data in an anonymous repository: https://anonymous.4open.science/r/MARS-GPS-DE55.
Read the full abstract ▾
All authors Md. Abu Bakor Siddique, Shahrin Hossain, Sadman Ahmed Siam, Syed Rifat Raiyan, Hasan Mahmud, Md Kamrul Hasan
Computation and Language · Cross-listed from cs.CV
PixelPrune: Pixel-Level Adaptive Visual Token Reduction via Predictive Coding
Nan Wang, Zhiwei Jin, Chen Chen et al. (+1 more)
Document understanding and GUI interaction are among the highest-value applications of
Vision-
Language Models (VLMs), yet they impose exceptionally heavy computational burden: fine-grained text and small UI elements demand high-resolution inputs that produce tens of thousands of visual
tokens. We observe that this cost is largely wasteful – across document and GUI
benchmarks, only 22–71% of image patches are pixel-unique, the rest being exact duplicates of another patch in the same image. We propose PixelPrune, which exploits this pixel-level redundancy through predictive-coding-based compression,
pruning redundant patches before the
Vision Transformer (ViT) encoder. Because it operates in pixel space prior to any neural computation, PixelPrune accelerates both the ViT encoder and the downstream LLM, covering the full inference pipeline. The method is training-free, requires no learnable parameters, and supports pixel-lossless compression ($\tau{=}0$) as well as controlled lossy compression ($\tau{>}0$). Experiments across three model scales and document and GUI
benchmarks show that PixelPrune maintains competitive task accuracy while delivering up to 4.2$\times$ inference speedup and 1.9$\times$ training acceleration. Code is available at https://github.com/OPPO-Mente-Lab/PixelPrune.
Read the full abstract ▾
All authors Nan Wang, Zhiwei Jin, Chen Chen, Haonan Lu
Computation and Language
StanceMoE: Mixture-of-Experts Architecture for Stance Detection
Abdullah Al Shafi, Md. Milon Islam, Sk. Imran Hossain et al. (+1 more)
Actor-level stance detection aims to determine an author expressed position toward specific geopolitical actors mentioned or implicated in a text. Although
transformer-based models have achieved relatively good performance in stance classification, they typically rely on unified representations that may not sufficiently capture heterogeneous linguistic signals, such as contrastive discourse structures, framing cues, and salient lexical indicators. This motivates the need for adaptive architectures that explicitly model diverse stance-expressive patterns. In this paper, we propose StanceMoE, a context-enhanced Mixture-of-Experts (MoE) architecture built upon a fine-tuned BERT encoder for actor-level stance detection. Our model integrates six expert modules designed to capture complementary linguistic signals, including global semantic orientation, salient lexical cues, clause-level focus, phrase-level patterns, framing indicators, and contrast-driven discourse shifts. A context-aware gating mechanism dynamically weights expert contributions, enabling adaptive routing based on input characteristics. Experiments are conducted on the StanceNakba 2026 Subtask A
dataset, comprising 1,401 annotated English texts where the target actor is implicit in the text. StanceMoE achieves a macro-F1 score of 94.26%, outperforming traditional baselines, and alternative BERT-based variants.
Read the full abstract ▾
All authors Abdullah Al Shafi, Md. Milon Islam, Sk. Imran Hossain, K. M. Azharul Hasan
Computation and Language
Agentic Tool Use in Large Language Models
Jinchao Hu, Meizhi Zhong, Kehai Chen et al. (+2 more)
Large
language models are increasingly being deployed as autonomous
agents yet their real world effectiveness depends on reliable tools for information
retrieval, computation and external action. Existing studies remain fragmented across tasks, tool types, and training settings, lacking a unified view of how tool-use methods differ and evolve. This paper organizes the literature into three paradigms: prompting as plug-and-play, supervised tool learning and reward-driven tool policy learning, analyzes their methods, strengths and failure modes, reviews the evaluation landscape and highlights key challenges, aiming to address this fragmentation and provide a more structured evolutionary view of agentic tool use.
Read the full abstract ▾
All authors Jinchao Hu, Meizhi Zhong, Kehai Chen, Xuefeng Bai, Min Zhang
Computation and Language · Cross-listed from cs.CV
LinguDistill: Recovering Linguistic Ability in Vision-Language Models via Selective Cross-Modal Distillation
Patrick Amadeus Irawan, Erland Hilman Fuadi, Shanu Kumar et al. (+2 more)
Turning a pretrained
language model (LM) into a
vision-
language model (VLM) through
multimodal fine-tuning often erodes its native
language ability, a form of catastrophic forgetting that shows up even on text-only tasks. This loss is hard to undo with further fine-tuning, and existing remedies add adapters or
alignment modules that increase architectural complexity and inference cost. We propose LinguDistill, an adapter-free knowledge
distillation method that uses the original frozen LM as the teacher during
multimodal post-training. To let a text-only teacher supervise
vision-conditioned outputs, we introduce layer-wise KV-cache sharing, which exposes the teacher to the student's
multimodal representations without changing either architecture. We then apply
distillation selectively, on
language-heavy data only, so the teacher restores linguistic ability while the student keeps its visual grounding on document and OCR tasks. LinguDistill recovers the
language and knowledge performance lost during
multimodal fine-tuning, matching the original VLM on average over text-only
benchmarks (ARC, HellaSwag) and exceeding it on ScienceQA, while keeping
vision-heavy performance close to standard fine-tuning. Since the teacher is dropped after training, the final model adds no parameters and no inference cost. More broadly, our results show that a model's own pre-adaptation backbone is a practical teacher for undoing forgetting, suggesting a simple recipe for keeping
language ability intact as models are extended to new modalities.
Read the full abstract ▾
All authors Patrick Amadeus Irawan, Erland Hilman Fuadi, Shanu Kumar, Alham Fikri Aji, Yova Kementchedjhieva
Computation and Language
Hemanth Kotaprolu, Kishan Maharaj, Raey Zhao et al. (+2 more)
Understanding
emotions in natural
language is inherently a multi-dimensional
reasoning problem, where multiple affective signals interact through context, interpersonal relations, and situational cues. However, most existing
emotion understanding
benchmarks rely on short texts and predefined
emotion labels, reducing this process to independent label prediction and ignoring the structured dependencies among
emotions. To address this limitation, we introduce Emotional Scenarios (EmoScene), a theory-grounded
benchmark of 4,731 contextrich scenarios annotated with an 8-dimensional
emotion vector derived from Plutchik's basic
emotions. Motivated by the observation that
emotions rarely occur independently, we further propose an
entanglement-aware
Bayesian inference framework that incorporates
emotion co-occurrence statistics to perform joint posterior inference over the
emotion vector. This lightweight post-processing does not require any parameter updates and improves the structural consistency of predictions, and yields overall gains of 2.24% Lexical Accuracy without any additional cost. EmoScene therefore provides a challenging
benchmark for studying multi-dimensional
emotion understanding and the limitations of current
language models.
Read the full abstract ▾
All authors Hemanth Kotaprolu, Kishan Maharaj, Raey Zhao, Abhijit Mishra, Pushpak Bhattacharyya
Computation and Language · Cross-listed from cs.LG
Routing-Free Mixture-of-Experts
Yilun Liu, Jinru Han, Sikuan Yan et al. (+2 more)
Standard Mixture-of-Experts (MoE) models rely on centralized routing mechanisms that introduce rigid inductive biases. We propose Routing-Free MoE which eliminates any hard-coded centralized designs including external routers, Softmax, Top-K and load balancing, instead encapsulating all activation functionalities within individual experts and directly optimized through continuous gradient flow, enabling each expert to determine its activation entirely on its own. We introduce a unified adaptive load-balancing framework to simultaneously optimize both expert-balancing and
token-balancing objectives through a configurable interpolation, allowing flexible and customizable resource allocation. Extensive experiments show that Routing-Free MoE can consistently outperform baselines with better scalability and robustness. We analyze its behavior in detail and offer insights that may facilitate future MoE design ad optimization.
Read the full abstract ▾
All authors Yilun Liu, Jinru Han, Sikuan Yan, Volker Tresp, Yunpu Ma
Computation and Language · Cross-listed from cs.CV
Multimodal Language Models Cannot Spot Spatial Inconsistencies
Om Khangaonkar, Hadi J. Rad and Hamed Pirsiavash
Spatial consistency is a fundamental property of the visual world and a key requirement for models that aim to understand physical reality. Despite recent advances,
multimodal large
language models (MLLMs) often struggle to reason about 3D geometry across multiple views. Rather than asking models to describe scene attributes, we introduce a more challenging task: given two views of the same scene, identify the object that violates 3D motion consistency. We propose a simple and scalable method for generating realistic, spatially inconsistent image pairs from multi-view scenes, enabling systematic evaluation of this capability. Our results show that state-of-the-art MLLMs significantly underperform human observers and exhibit substantial variability across different scene attributes, revealing a fragile and incomplete understanding of 3D structure. We hope our findings underscore the need for approaches that develop a more deeply grounded understanding of the physical world.
Read the full abstract ▾
Computation and Language
Valency Classification of Mapudungun Verbal Roots. Established by the language's own morphotactics
Andrés Chandía
In the previous work, a lexical (re)categorisation – or confirmation of the given category – of roots identified as verbal was undertaken to determine their original category accurately. Building on this, the present paper offers an account of the valency classification of those Mapudungun roots confirmed to be verbal, using the
language's own morphotactics; specifically, by examining the permissible and restricted combinations of various suffixes with roots or verbal stems in the Mapuche verb form. As with all work conducted thus far, the results presented here aim to improve the morphological analyser (Dungupeyum) with all verified findings incorporated into the system. From a theoretical perspective, we also hope to contribute to the recognition and understanding of issues related to the valency of Mapuche verb forms.
Read the full abstract ▾
Computation and Language
From Early Encoding to Late Suppression: Interpreting LLMs on Character Counting Tasks
Ayan Datta, Mounika Marreddy, Alexander Mehler et al. (+2 more)
Large
language models (LLMs) exhibit failures on elementary symbolic tasks such as character counting in a word, despite excelling on complex
benchmarks. Although this limitation has been noted, the internal reasons remain unclear. We use character counting (e.g., "How many p's are in apple?") as a minimal, controlled probe that isolates
token-level
reasoning from higher-level confounds. Using this setting, we uncover a consistent phenomenon across modern architectures, including LLaMA, Qwen, and Gemma: models often compute the correct answer internally yet fail to express it at the output layer. Through mechanistic analysis combining probing classifiers, activation patching, logit lens analysis, and
attention head tracing, we show that character-level information is encoded in early and mid-layer representations. However, this information is attenuated by a small set of components in later layers, especially the penultimate and final layer MLP. We identify these components as negative
circuits: subnetworks that downweight correct signals in favor of higher-probability but incorrect outputs. Our results lead to two contributions. First, we show that symbolic
reasoning failures in LLMs are not due to missing representations or insufficient scale, but arise from structured interference within the model's computation graph. This explains why such errors persist and can worsen under scaling and instruction tuning. Second, we provide evidence that LLM forward passes implement a form of competitive decoding, in which correct and incorrect hypotheses coexist and are dynamically reweighted, with final outputs determined by suppression as much as by amplification. These findings carry implications for interpretability and robustness: simple symbolic
reasoning exposes weaknesses in modern LLMs, underscoring need for design strategies that ensure information is encoded and reliably used.
Read the full abstract ▾
All authors Ayan Datta, Mounika Marreddy, Alexander Mehler, Zhixue Zhao, Radhika Mamidi
Computation and Language
From Baselines to Preferences: A Comparative Study of LoRA/QLoRA and Preference Optimization for Mental Health Text Classification
Mihael Arcan
Mental health text classification has rapidly adopted modern adaptation methods, yet practical guidance on which optimization strategy to use, when, and why remains limited. This paper presents a systematic comparative study of optimization pathways for a joint mental-health classification task, moving from strong vanilla baselines to progressively more specialized techniques. We first establish classical and encoder references, then examine parameter-efficient supervised fine-tuning with LoRA/QLoRA under multiple objective and optimization settings, and finally evaluate preference-based optimization with DPO, ORPO, and KTO, including class-rebalanced training. Rather than emphasizing a single headline score, we focus on methodological insight: how performance changes with objective formulation, adapter choice, optimizer behavior, context windowing, and class-balance intervention. The results show that optimization effects are highly method-dependent: some approaches deliver stable, transferable gains, while others are sensitive to configuration and data balance. Preference optimization, in particular, exhibits large variation across objectives, indicating that method selection is more consequential than simply adding a preference-training stage. The central contribution is a clear optimization narrative for mental health NLP: start from transparent baselines, apply controlled tuning, and use preference optimization selectively where its gains are demonstrable. This provides a reproducible and practically grounded framework for choosing effective training strategies beyond architecture choice alone.
Read the full abstract ▾
Computation and Language
Stochastic Attention: Connectome-Inspired Randomized Routing for Expressive Linear-Time Attention
Zehao Jin and Yanan Sui
The whole-
brain connectome of a fruit fly comprises over 130K
neurons connected with a probability of merely 0.02%, yet achieves an average shortest path of only 4.4 hops. Despite being highly structured at the
circuit level, the network's long-range connections are broadly distributed across
brain regions, functioning as
stochastic shortcuts that enable efficient global communication. Inspired by this observation, we propose
Stochastic Attention (SA), a drop-in enhancement for sliding-window
attention (SWA) that applies a random permutation to the
token sequence before windowed
attention and restores the original order afterward. This transforms the fixed local window into a
stochastic global one within the same $O(nw)$ per-layer budget. Through depth, independently sampled permutations yield exponentially growing receptive fields, achieving full sequence coverage in $O(\log_w n)$ layers versus $O(n/w)$ for SWA. We validate SA in two settings: pre-training
language models from scratch, where a gated SA + SWA combination achieves the best average zero-shot accuracy, and training-free inference on Qwen3-8B and Qwen3-30B-A3B, where SA consistently outperforms SWA and matches or exceeds Mixture of Block
Attention at comparable compute budgets. These results suggest that connectome-inspired
stochastic routing is a practical primitive for improving the expressivity of efficient
attention, complementary to existing linear and sparse approaches.
Read the full abstract ▾
Computation and Language
LangMARL: Natural Language Multi-Agent Reinforcement Learning
Huaiyuan Yao, Longchao Da, Xiaoou Liu et al. (+3 more)
Large
language model (LLM)
agents struggle to autonomously evolve coordination strategies in dynamic environments, largely because coarse global outcomes obscure the causal signals needed for local policy refinement. We identify this bottleneck as a multi-
agent credit assignment problem, which has long been studied in classical multi-
agent reinforcement learning (MARL) but remains underaddressed in LLM-based systems. Building on this observation, we propose LangMARL, a framework that brings credit assignment and policy gradient
evolution from cooperative MARL into the
language space. LangMARL introduces
agent-level
language credit assignment, pioneers gradient
evolution in
language space for policy improvement, and summarizes task-relevant causal relations from replayed trajectories to provide dense feedback and improve convergence under sparse rewards. Extensive experiments across diverse cooperative multi-
agent tasks demonstrate improved sample efficiency, interpretability, and strong generalization.
Read the full abstract ▾
All authors Huaiyuan Yao, Longchao Da, Xiaoou Liu, Charles Fleming, Tianlong Chen, Hua Wei
Computation and Language
To Memorize or to Retrieve: Scaling the Interaction Between Pretraining and Retrieval
Karan Singh, Michael Yu, Varun Gangal et al. (+4 more)
Retrieval-augmented generation (RAG) improves
language model (LM) performance by providing relevant context at test time for knowledge-intensive situations. In this work, we systematically study the trade-off between pretraining and
retrieval by training OLMo-2-based LMs ranging from 30M to 3B parameters on up to 100B DCLM
tokens, while varying pretraining data scale,
retrieval store size, and
retrieval store source (pretraining vs. new data) across
reasoning, scientific QA, and open-domain QA
benchmarks. We find that
retrieval gains depend on model capacity and pretraining exposure and are strongly front-loaded, with a median 91% of the largest observed improvement realized by one
retrieval token per model parameter. However, the interaction is objective-dependent: smaller models gain more in gold-answer perplexity, whereas larger, more-pretrained models gain more in accuracy.
Retrieval from previously seen data also preserves most of the held-out
retrieval gain.
Retrieval is therefore a task-, regime-, and metric-dependent complement to parametric learning whose value also depends on datastore size and information novelty. Overall, this motivates the explicit partitioning of data between internalization and external access for LM design.
Read the full abstract ▾
All authors Karan Singh, Michael Yu, Varun Gangal, Zhuofu Tao, Sachin Kumar, Emmy Liu, Steven Y. Feng
Computation and Language
AfrIFact: Cultural Information Retrieval, Evidence Extraction and Fact Checking for African Languages
Israel Abebe Azime, Jesujoba Oluwadara Alabi, Crystina Zhang et al. (+16 more)
Assessing the veracity of a claim made online is a complex and important task with real-world implications. When these claims are directed at communities with limited access to information and the content concerns issues such as healthcare and culture, the consequences intensify, especially in low-resource
languages. In this work, we introduce AfrIFact, a
dataset that covers the necessary steps for automatic fact-checking (i.e., information
retrieval, evidence extraction, and fact checking), in ten African
languages and English. Our evaluation results show that even the best
embedding models lack cross-lingual
retrieval capabilities, and that cultural and news documents are easier to retrieve than healthcare-domain documents, both in large corpora and in single documents. We show that LLMs lack robust
multilingual fact-verification capabilities in African
languages, while few-shot prompting improves performance by up to 43% in AfriqueQwen-14B, and task-specific fine-tuning further improves fact-checking accuracy by up to 26%. These findings, along with our release of the AfrIFact
dataset, encourage work on low-resource information
retrieval, evidence
retrieval, and fact checking.
Read the full abstract ▾
All authors Israel Abebe Azime, Jesujoba Oluwadara Alabi, Crystina Zhang, Iffat Maab, Atnafu Lambebo Tonja, Tadesse Destaw Belay, Folasade Peace Alabi, Salomey Osei, Saminu Mohammad Aliyu, Nkechinyere Faith Aguobi, Bontu Fufa Balcha, Blessing Kudzaishe Sibanda, Davis David, Mouhamadane Mboup, Daud Abolade, Neo Putini, Philipp Slusallek, David Ifeoluwa Adelani, Dietrich Klakow
Computation and Language · Cross-listed from cs.LG
Learning to Hint for Reinforcement Learning
Yu Xia, Canwen Xu, Zhewei Yao et al. (+2 more)
Group Relative Policy Optimization (GRPO) is widely used for reinforcement learning with verifiable rewards, but it often suffers from advantage collapse: when all rollouts in a group receive the same reward, the group yields zero relative advantage and thus no learning signal. For example, if a question is too hard for the reasoner, all sampled rollouts can be incorrect and receive zero reward. Recent work addresses this issue by adding hints or auxiliary scaffolds to such hard questions so that the reasoner produces mixed outcomes and recovers a non-zero update. However, existing hints are usually fixed rather than adapted to the current reasoner, and a hint that creates learning signal under the hinted input does not necessarily improve the no-hint policy used at test time. To this end, we propose Hint Learning for Reinforcement Learning (HiLL), a framework that jointly trains a hinter policy and a reasoner policy during RL. For each hard question, the hinter generates hints online conditioned on the current reasoner's incorrect rollout, allowing hint generation to adapt to the reasoner's evolving errors. We further introduce hint reliance, which measures how strongly correct hinted trajectories depend on the hint. We derive a transferability result showing that lower hint reliance implies stronger transfer from hinted success to no-hint success, and we use this result to define a transfer-weighted reward for training the hinter. Therefore, HiLL favors hints that not only recover informative GRPO groups, but also produce signals that are more likely to improve the original no-hint policy. Experiments across multiple
benchmarks show that HiLL consistently outperforms GRPO and prior hint-based baselines, demonstrating the value of adaptive and transfer-aware hint learning for RL. The code is available at https://github.com/Andree-9/HiLL.
Read the full abstract ▾
All authors Yu Xia, Canwen Xu, Zhewei Yao, Julian McAuley, Yuxiong He
Computation and Language
OmniVoice: Towards Omnilingual Zero-Shot Text-to-Speech with Diffusion Language Models
Han Zhu, Lingxuan Ye, Wei Kang et al. (+7 more)
We present OmniVoice, a massively
multilingual zero-shot text-to-
speech (TTS) model that scales to over 600
languages. At its core is a novel
diffusion language model-style discrete non-autoregressive (NAR) architecture. Unlike conventional discrete NAR models that suffer from performance bottlenecks in complex two-stage (text-to-semantic-to-acoustic) pipelines, OmniVoice directly maps text to multi-codebook acoustic
tokens. This simplified approach is facilitated by two key technical innovations: (1) a full-codebook random masking strategy for efficient training, and (2) initialization from a pre-trained LLM to ensure superior intelligibility. By leveraging a 581k-hour
multilingual dataset curated entirely from open-source data, OmniVoice achieves the broadest
language coverage to date and delivers state-of-the-art performance across Chinese, English, and diverse
multilingual benchmarks. Our code and pre-trained models are publicly available at https://github.com/k2-fsa/OmniVoice.
Read the full abstract ▾
All authors Han Zhu, Lingxuan Ye, Wei Kang, Zengwei Yao, Liyong Guo, Fangjun Kuang, Zhifeng Han, Weiji Zhuang, Long Lin, Daniel Povey
Computation and Language
Common TF-IDF variants arise as key components in the test statistic of a penalized likelihood-ratio test for word burstiness
Zeyad Ahmed, Paul Sheridan, Michael McIsaac et al. (+1 more)
TF-IDF is a classical formula that is widely used for identifying important terms within documents. We show that TF-IDF-like scores arise naturally from the test statistic of a penalized likelihood-ratio test setup capturing word burstiness (also known as word over-dispersion). In our framework, the alternative hypothesis captures word burstiness by modeling a collection of documents according to a family of beta-binomial distributions with a gamma penalty term on the precision parameter. In contrast, the null hypothesis assumes that words are binomially distributed in collection documents, a modeling approach that fails to account for word burstiness. We find that a term-weighting scheme given rise to by this test statistic performs comparably to TF-IDF on document classification tasks. This paper provides insights into TF-IDF from a statistical perspective and underscores the potential of hypothesis testing frameworks for advancing term-weighting scheme development.
Read the full abstract ▾
All authors Zeyad Ahmed, Paul Sheridan, Michael McIsaac, Aitazaz A. Farooque
Computation and Language
TRIMS: Trajectory-Ranked Instruction Masked Supervision for Diffusion Language Models
Lingjie Chen, Ruizhong Qiu, Yuyu Fan et al. (+2 more)
Diffusion language models (DLMs) offer a promising path toward low-latency generation through parallel decoding, but their practical efficiency depends heavily on the decoding trajectory. In practice, this advantage often fails to fully materialize because standard training does not provide explicit supervision over
token reveal order, creating a train-inference mismatch that leads to suboptimal decoding behavior. We propose Trajectory-Ranked Instruction Masked Supervision (TRIMS), a simple trajectory-guided supervised fine-tuning framework that injects trajectory supervision into standard Masked
Diffusion Language Model (MDLM) training with minimal overhead. Instead of relying on costly DLM-based
distillation, TRIMS uses lightweight signals from an autoregressive teacher to guide a trajectory-aware masking strategy, encouraging the model to learn more effective decoding orders. Experiments on LLaDA and Dream across math and coding
benchmarks show that TRIMS significantly improves the accuracy-parallelism trade-off over both standard MDLM training and train-free acceleration baselines, while achieving competitive performance with prior
distillation-based approaches at substantially lower training cost. Further analysis shows that TRIMS leads to better decoding trajectories, validating the effectiveness of trajectory-guided supervision for DLMs.
Read the full abstract ▾
All authors Lingjie Chen, Ruizhong Qiu, Yuyu Fan, Yanjun Zhao, Hanghang Tong
Computation and Language · Cross-listed from cs.LG
Mingyang Song and Mao Zheng
On-Policy
Distillation (OPD) trains large
language models with teacher guidance on student-generated trajectories, addressing the mismatch between fixed demonstrations and the histories encountered during deployment. As its use expands across
language model training, different formulations of supervision and optimization make it difficult to compare methods and understand when they are effective. This survey develops a unified view of OPD through the interplay between student-generated states, teacher information, and policy updates. This framework connects the foundations of
distillation with self-
distillation, multi-teacher integration, reinforcement learning, and interactive
agents. We examine how these settings change the supervision available to the student and relate the resulting design choices to capability transfer, training efficiency, and deployment. Our synthesis centers on the conditions under which guidance remains reliable and learnable as the student evolves, considering both immediate performance and the capacity for subsequent learning. We conclude by discussing open challenges in adapting supervision and allocating computation to support effective learning.
Read the full abstract ▾
Computation and Language
English to Central Kurdish Speech Translation: Corpus Creation, Evaluation, and Orthographic Standardization
Mohammad Mohammadamini, Daban Q. Jaff, Josep Crego et al. (+2 more)
We present KUTED, a
speech-to-text translation (S2TT)
dataset for Central Kurdish, derived from TED and TEDx talks. The corpus comprises 91,000 sentence pairs, including 170 hours of English audio, 1.65 million English
tokens, and 1.40 million Central Kurdish
tokens. We evaluate KUTED on the S2TT task and find that orthographic variation significantly degrades Kurdish translation performance, producing nonstandard outputs. To address this, we propose a systematic text standardization approach that yields substantial performance gains and more consistent translations. On a test set separated from TED talks, a fine-tuned Seamless model achieves 15.18 BLEU, and we improve Seamless baseline by 3.0 BLEU on the FLEURS
benchmark. We also train a
Transformer model from scratch and evaluate a cascaded system that combines Seamless (ASR) with NLLB (MT).
Read the full abstract ▾
All authors Mohammad Mohammadamini, Daban Q. Jaff, Josep Crego, Marie Tahon, Antoine Laurent
Computation and Language
Speech LLMs are Contextual Reasoning Transcribers
Keqi Deng, Ruchao Fan, Bo Ren et al. (+2 more)
Despite extensions to
speech inputs, effectively leveraging the rich knowledge and contextual understanding of large
language models (LLMs) in automatic
speech recognition (ASR) remains non-trivial, as the task primarily involves direct
speech-to-text mapping. To address this, this paper proposes chain-of-thought ASR (CoT-ASR), which constructs a
reasoning chain that enables LLMs to first analyze the input
speech and generate contextual analysis, thereby fully exploiting their generative capabilities. With this contextual
reasoning, CoT-ASR then performs more informed
speech recognition and completes both
reasoning and transcription in a single pass. Moreover, CoT-ASR naturally supports user-guided transcription: while designed to self-generate
reasoning, it can also seamlessly incorporate user-provided context to guide transcription, further extending ASR functionality. To reduce the modality gap, this paper introduces a CTC-guided Modality Adapter, which uses CTC non-blank
token probabilities to weight LLM
embeddings, efficiently aligning
speech encoder outputs with the LLM's textual latent space. Experiments show that, compared to standard LLM-based ASR, CoT-ASR achieves a relative reduction of 8.7% in word error rate (WER) and 16.9% in entity error rate (EER).
Read the full abstract ▾
All authors Keqi Deng, Ruchao Fan, Bo Ren, Yiming Wang, Jinyu Li
Computation and Language
More Human, More Efficient: Aligning Annotations with Quantized SLMs
Jiayu Wang and Junyoung Lee
As Large
Language Model (LLM) capabilities advance, the demand for high-quality annotation of exponentially increasing text corpora has outpaced human capacity, leading to the widespread adoption of LLMs in automatic evaluation and annotation. However, proprietary LLMs often exhibit systematic biases that diverge from human expert consensus, lacks reproducibility, and raises data
privacy concerns. Our work examines the viability of finetuning a quantized Small
Language Model of 1.7B parameter size on limited human-annotated data to serve as a highly aligned, deterministic evaluator and annotator. By implementing a custom, multi-dimensional rubric framework and simple augmentation and regularization techniques, the proposed approach achieves higher inter-annotator agreement (0.23 points increase in Krippendorff's $\alpha$) than the best performing state-of-the-art proprietary LLM. We also demonstrate the generalizability of the proposed training pipeline on a separate
emotion classification task. The results show that task-specific
alignment and efficient 4-bit quantized fine-tuning provide superior open-source alternative to using proprietary models for evaluation and annotation. Our finetuning approach is publicly available at https://github.com/jylee-k/slm-judge.
Read the full abstract ▾
Computation and Language
A Japanese Benchmark for Evaluating Social Bias in Reasoning Based on Attribution Theory
Taihei Shiotani, Masahiro Kaneko and Naoaki Okazaki
In enhancing the
fairness of Large
Language Models (LLMs), evaluating social biases rooted in the cultural contexts of specific linguistic regions is essential. However, most existing Japanese
benchmarks heavily rely on translating English data, which does not necessarily provide an evaluation suitable for Japanese culture. Furthermore, they only evaluate bias in the conclusion, failing to capture biases lurking in the
reasoning. In this study, based on attribution theory in social psychology, we constructed a new
dataset, “JUBAKU-v2,” which evaluates the bias in attributing behaviors to in-groups and out-groups within
reasoning while fixing the conclusion. This
dataset consists of 216 examples reflecting cultural biases specific to Japan. Experimental results verified that it can detect performance differences across models more sensitively than existing
benchmarks.
Read the full abstract ▾
Computation and Language · Cross-listed from cs.AI
Ontology-Constrained Neural Reasoning in Enterprise Agentic Systems: A Neurosymbolic Architecture for Domain-Grounded AI Agents
Thanh Luong Tuan and Abhijit Sanyal
Enterprise adoption of Large
Language Models (LLMs) is constrained by
hallucination, domain drift, and the inability to enforce regulatory compliance at the
reasoning level. We present a neurosymbolic architecture implemented within the Foundation AgenticOS (FAOS) platform that addresses these limitations through ontology-constrained neural
reasoning. We introduce a three-layer ontological framework–Role, Domain, and Interaction ontologies–grounding LLM-based enterprise
agents. We formalize asymmetric neurosymbolic coupling: current enterprise systems constrain
agent inputs (context assembly, tool discovery, governance thresholds) but not outputs, and we propose mechanisms extending this coupling to output-side validation (response checking,
reasoning verification, compliance enforcement). A controlled experiment (1,800 runs across five industries and three LLMs: Claude Sonnet 4, Qwen 2.5 72B, Gemma 4 26B) finds ontology-coupled
agents significantly outperform ungrounded
agents on Metric Accuracy (p < .001) and Role Consistency (p < .001) across all three models with large effect sizes (Kendall's W = .46-.64). Improvements are greatest where LLM parametric knowledge is weakest–particularly in Vietnam-localized domains, where ontology lift is 2x that of English domains. Contributions: (1) a formal three-layer enterprise ontology model; (2) a taxonomy of neurosymbolic coupling patterns; (3) ontology-constrained tool discovery via SQL-pushdown scoring; (4) a proposed framework for output-side ontological validation; (5) empirical evidence for the inverse parametric knowledge effect–ontological grounding value is inversely proportional to LLM training-data coverage of the domain; (6) cross-model replication establishing model-independence; (7) a production system serving 22 industry verticals with 650+
agents.
Read the full abstract ▾
Computation and Language
Optimsyn: Influence-Guided Rubrics Optimization for Synthetic Data Generation
Zhiting Fan, Ruizhe Chen, Tianxiang Hu et al. (+7 more)
Large
language models (LLMs) achieve strong downstream performance largely due to abundant supervised fine-tuning (SFT) data. However, high-quality SFT data in knowledge-intensive domains such as humanities, social sciences, medicine, law, and finance is scarce because expert curation is expensive,
privacy constraints are strict, and label consistency is hard to ensure. Recent work uses synthetic data, typically by prompting a generator over domain documents and filtering outputs with handcrafted rubrics. Yet rubric design is expert-dependent, transfers poorly across domains, and is often optimized through a brittle heuristic loop of writing rubrics, synthesizing data, training, inspecting results, and manually guessing revisions. This process lacks reliable quantitative feedback about how a rubric affects downstream performance. We propose evaluating synthetic data by its training utility on the target model and using this signal to guide data generation. Inspired by influence estimation, we adopt an optimizer-aware
estimator that uses gradient information to quantify each synthetic sample's contribution to a target model's objective on specific tasks. Our analysis shows that even when synthetic and real samples are close in
embedding space, their influence on learning can differ substantially. Based on this insight, we propose an optimization-based framework that adapts rubrics using target-model feedback. We provide lightweight guiding text and use a rubric-specialized model to generate task-conditioned rubrics. Influence score is used as the reward to optimize the rubric generator with reinforcement learning. Experiments across domains, target models, and data generators show consistent improvements and strong generalization without task-specific tuning.
Read the full abstract ▾
All authors Zhiting Fan, Ruizhe Chen, Tianxiang Hu, Ru Peng, Zenan Huang, Haokai Xu, Yixin Chen, Jian Wu, Junbo Zhao, Zuozhu Liu
Computation and Language · Cross-listed from cs.LG
MF-QAT: Multi-Format Quantization-Aware Training for Elastic Inference
Zifei Xu, Sayeh Sharify and Hesham Mostafa
Quantization-aware training (QAT) is typically performed for a single target numeric format, while practical deployments often need to choose numerical precision at inference time based on hardware support or runtime constraints. We study multi-format QAT, where a single model is trained to be robust across multiple quantization formats. We find that multi-format QAT can match single-format QAT at each target precision, yielding one model that performs well overall across different formats, even formats that were not seen during training. To enable practical deployment, we propose the Slice-and-Scale conversion procedure for both MXINT and MXFP that converts a high-precision representation into lower-precision formats without re-training. Building on this, we introduce a pipeline that (i) trains a model with multi-format QAT, (ii) stores a single anchor format checkpoint (MXINT8/MXFP8), and (iii) allows on-the-fly conversion to lower MXINT or MXFP formats at runtime with negligible-or no-additional accuracy degradation. Together, these components provide a practical path to elastic precision scaling and allow selecting the runtime format at inference time across diverse deployment targets.
Read the full abstract ▾
Computation and Language
Adapting Text LLMs to Speech via Multimodal Depth Up-Scaling
Kazuki Yano, Jun Suzuki and Shinji Watanabe
Adapting pre-trained text Large
Language Models (LLMs) into
Speech Language Models (
Speech LMs) via continual pretraining on
speech data is promising, but often degrades the original text capabilities. We propose
Multimodal Depth Upscaling, an extension of an emerging strategy in continual LLM pre-training, where new
transformer layers are inserted into a frozen text LLM and only the added layers are trained on
speech data. Experiments with SmolLM2-360M and SmolLM2-1.7B on 48k hours of English Automatic
Speech Recognition (ASR) data show that depth up-scaling achieves ASR comparable to full fine-tuning while causing far less text degradation than both full fine-tuning and Low-Rank Adaptation (LoRA). We further show that incorporating E-Branchformer, an architecture designed for
speech recognition, as the inserted layers achieves ASR that matches or surpasses full fine-tuning on the larger model while reducing text degradation by over 75% with 60% fewer trainable parameters.
Read the full abstract ▾
Computation and Language · Cross-listed from cs.AI
Logarithmic Scores, Power-Law Discoveries: Disentangling Measurement from Coverage in Agent-Based Evaluation
HyunJoon Jung and William Na
LLM-based
agent judges are an emerging approach to evaluating conversational AI, yet a fundamental uncertainty remains: can we trust their assessments, and if so, how many are needed? Through 960 sessions with two model pairs across 15 tasks, we show that persona-based
agent judges produce evaluations indistinguishable from human raters in a Turing-style validation. We then identify a score-coverage dissociation: quality scores improve logarithmically with panel size, while unique issue discoveries follow a sublinear power law-both exhibit diminishing returns, but scores saturate roughly twice as fast as discoveries. We hypothesize this reflects a power law distribution of the finding space: critical issues are discovered first by small panels, while corner cases require progressively larger panels, analogous to species accumulation curves in ecology. The mechanism traces to ensemble diversity-Big Five personality conditioning makes
agents probe different quality dimensions, with expert judges acting as
adversarial probes that push discovery into the tail of the finding distribution. A controlled ablation confirms that structured persona conditioning, not simple prompting, is required to produce these scaling properties.
Read the full abstract ▾
Computation and Language · Cross-listed from cs.HC
Not My Truce: Personality Differences in AI-Mediated Workplace Negotiation
Veda Duddu, Jash Rajesh Parekh, Andy Mao et al. (+4 more)
AI-driven conversational coaching is increasingly used to support workplace negotiation, yet prior work assumes uniform effectiveness across users. We challenge this assumption by examining how individual differences, particularly personality traits, moderate coaching outcomes. We conducted a between-subjects experiment (N=267) comparing theory-driven AI (Trucey), general-purpose AI (Control-AI), and a traditional negotiation handbook (Control-NoAI). Participants were clustered into three profiles – resilient, overcontrolled, and undercontrolled – based on the Big-Five personality traits and ARC typology. Resilient workers achieved broad psychological gains primarily from the handbook, overcontrolled workers showed outcome-specific improvements with theory-driven AI, and undercontrolled workers exhibited minimal effects despite engaging with the frameworks. These patterns suggest personality as a predictor of readiness beyond stage-based tailoring: vulnerable users benefit from targeted rather than comprehensive interventions. The study advances understanding of personality-determined intervention prerequisites and highlights design implications for adaptive AI coaching systems that align support intensity with individual readiness, rather than assuming universal effectiveness.
Read the full abstract ▾
All authors Veda Duddu, Jash Rajesh Parekh, Andy Mao, Hanyi Min, Ziang Xiao, Vedant Das Swain, Koustuv Saha