Peer Revenue Fully funded

Papers

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

Computer Science Computation and Language

cs.CL Watch Computation and Language

Monday 25 May

Computation and Language

Can LLMs Time Travel? Enhancing Temporal Consistency in Legal Agentic Search through Reinforcement Learning

Wei Fan, Yining Zhou, Mufan Zhang et al. (+8 more)

While large language models (LLMs) augmented with agentic search capabilities show promise for legal reasoning, they overlook a fundamental constraint that applicable law must match the temporal context of each case, as retroactive application of statutes violates core legal principles and leads to erroneous conclusions. Our observations reveal that current legal LLMs suffer from temporal bias anchored to their training cutoff, while search agents rarely incorporate temporal constraints into queries, and that web search alone cannot provide the precise statute and precedent citations that legal reasoning demands. To address these challenges, we propose LegalSearch-R1, an end-to-end reinforcement learning framework that pairs local statute RAG for precise article matching with online web search for broader legal knowledge, trained on temporally-indexed data spanning multiple amendment periods to enforce temporal consistency. Extensive experiments on our benchmark covering 13 legal tasks demonstrate that our 7B-parameter agent outperforms state-of-the-art deep research frameworks and specialized legal LLMs by 12.9% to 29.8%, surpasses baselines by 57.7% to 80.3% on temporal consistency, and exhibits robust out-of-domain generalization. The code and data are available at https://github.com/AlexFanw/LegalSearch-R1.
Read the full abstract

All authors Wei Fan, Yining Zhou, Mufan Zhang, Yanbing Weng, Yiran HU, Tianshi Zheng, Baixuan Xu, Chunyang Li, Jianhui Yang, Haoran Li, Yangqiu Song

Computation and Language

Universal Activation Verbalizer: A Unified Framework for Cross-Model Activation Explanation

Haiyan Zhao, Zirui He, Guanchu Wang et al. (+3 more)

Activation verbalization explains hidden representations in natural language, but existing methods are mostly limited to self-explanation, where each model explains only its own activations. We introduce Universal Activation Verbalizer (UAV), a framework that uses a shared decoder to explain activations from heterogeneous donor models. UAV learns a lightweight adapter that converts donor activations into soft tokens in decoder's embedding space, and further supports adapter-only transfer by reusing a frozen decoder-side LoRA while training only a new adapter for another donor. Across classification, fact retrieval, and gist summarization, UAV remains competitive with strong self-explanation baselines while enabling cross-model verbalization across model families and scales. Ablations show that decoder-side tuning mainly improves task behavior, whereas the adapter provides the activation-grounded factual and semantic information needed for faithful explanations. Code and data are available at https://github.com/hy-zhao23/ActExp.
Read the full abstract

All authors Haiyan Zhao, Zirui He, Guanchu Wang, Ali Payani, Yingcong Li, Mengnan Du

Computation and Language

Causal Tongue-Tie: LLMs Can Encode Causal Direction, But Their Yes/No Outputs Fail to Express

Ziyi Ding and Xiao-Ping Zhang

We find a mismatch between what large language models encode about a causal question and what they answer. On anti-commonsense CLadder items, a fixed linear probe recovers the evidence-supported answer from the model's hidden state (accuracy approximately 0.97), while the spoken Yes/No reverts to the commonsense one (accuracy approximately 0.5). We call this approximately +0.5 gap Causal Tongue-Tie: a wrong Yes/No decomposes into two separable failure modes: no internal signal versus a signal the verbal interface cannot say. The implication cuts both ways for output-only causal benchmarks: a benchmark "correct" need not mean the model has understood, and a benchmark "wrong" need not mean it cannot. Sweeping claims about whether LLMs can do causal reasoning, drawn from a single accuracy number, deserve a second look.
Read the full abstract

Computation and Language

Mitigating Provenance-Role Collapse in Long-Term Agents via Typed Memory Representation

Zhengda Jin, Bingbing Wang, Jing Li et al. (+2 more)

Long-term memory is essential for persistent LLM agents, yet prevailing architectures store historical interactions as unstructured, flat text. This unconstrained storage induces provenance-role collapse, a critical failure mode where agents suffer from source-monitoring errors. To resolve this cognitive vulnerability at the architectural level, we propose MemIR, a typed Memory Intermediate Representation that operationalizes source monitoring as a structural constraint. MemIR writes long-term memory into grounded atoms that separate raw evidence, retrieval cues, and truth-bearing claims, with factual authorization restricted to supported claim atoms. It then applies multi-route atomic projection and provenance-scoped utilization to transform heterogeneous retrieval hits into claim-centered candidate bundles and a normalized fact interface for answer generation. Experiments on LoCoMo and BEAM-100K demonstrate that MemIR consistently outperforms existing memory baselines, especially on tasks requiring source tracking, temporal grounding, and aggregation of fragmented evidence.
Read the full abstract

All authors Zhengda Jin, Bingbing Wang, Jing Li, Ruifeng Xu, Min Zhang

Computation and Language · Cross-listed from cs.LG

When Self-Belief Misleads: Active Label Acquisition for Reinforcement Learning with Verifiable Rewards

Li Wang, Xiaodong Lu, Xiaohan Wang et al. (+5 more)

Large Language Models (LLMs) have achieved remarkable advancements in reasoning capabilities empowered by Reinforcement Learning with Verifiable Rewards (RLVR). Nonetheless, RLVR intrinsically relies on ground-truth labels for reward computation, the acquisition of which is often prohibitively expensive in real-world scenarios. While unsupervised RLVR paradigms attempt to circumvent this by training on pseudo-labels, they are notoriously susceptible to training collapse. Moreover, different samples often exhibit varying annotation values. In this paper, we propose Reinforcement Learning with Active Verifiable Rewards (RLAVR), which actively acquires ground-truth labels for a small set of selected samples and integrates them with pseudo-labels, thereby stabilizing training dynamics and improving performance under limited annotation budgets. To identify valuable samples, we propose the Corrective Advantage Gap (CAG) metric and analyze the sample-level supervision value. Building on this, we introduce Correction-Aware Reliability Estimation for RLAVR (CARE), which translates the oracle CAG criterion into a practical pre-query acquisition policy to substantially improve training stability. Extensive experiments across diverse domains, model families, and model scales demonstrate the effectiveness and generality of our approach. Our code is available at https://github.com/Lumina04/CARE.
Read the full abstract

All authors Li Wang, Xiaodong Lu, Xiaohan Wang, Yikun Ban, Jiajun Chai, Wei Lin, Tianhao Peng, Guojun Yin

Computation and Language

TIAR: Trajectory-Informed Advantage Reweighting for LLM Abstention Learning

Muyu Pan, Shu Zhao, Nan Zhang et al. (+4 more)

This paper investigates large language model (LLM) abstention learning, specifically using ternary reward, which incentivize truthfulness in large language models. This paper extends that idea by moving from a ternary reward to a Trajectory-Informed advantage reweighting, dynamically re-weights the abstention reward during Group Relative Policy Optimization (GRPO) training. The objective of this work focuses on abstention learning instead of improving truthfulness, serving as an exploration into hallucination reduction. The novelty of this paper lies in methodological innovation, advantage re-weighting, and benchmark selection. Leveraging GRPO's multiple trajectories as a natural abstention signal, this method uses a reward signal to explore knowledge boundaries and encourage consistency. By demonstrating that trajectories can be used as a confidence indicator of the policy relative to the query, they are then used to dynamically calculate the abstention advantage. AbstentionBench is used as the evaluation benchmark, as this work aims to contribute to the field of abstention learning. All datasets on the benchmark were tested against this method and various baselines. Empirical results demonstrate that TIAR achieves state-of-the-art abstention F1 scores across five of six evaluation categories, outperforming the static ternary baseline on 17 of 31 benchmark datasets while fully preserving baseline accuracy.
Read the full abstract

All authors Muyu Pan, Shu Zhao, Nan Zhang, Philip Shin, Varun Parekh, Vijaykrishnan Narayanan, Rui Zhang

Computation and Language

On the Limits of Model Merging for Multilinguality in Pre-Training

Seth Aycock, Fedor Vitiugin, Aleksandr Umnov et al. (+2 more)

Endowing models with consistent multilingual performance can be achieved by mixing pre-training data, or post-training approaches such as language-specific model merging. In this work, we test whether merging can be applied to monolingually pre-trained models. We conduct a controlled study on the efficacy of mixed, merged, and monolingual pre-training setups. We find that while monolingual pre-training results in strong in-language performance, merging any combination of monolingual models leads to performance collapse due to interference. Our analysis suggests representational similarity is a prerequisite for model merging. We therefore conclude that the flexibility of merging in fine-tuning does not extend trivially to language-specific pre-training.
Read the full abstract

All authors Seth Aycock, Fedor Vitiugin, Aleksandr Umnov, Christof Monz, Khalil Sima'an

Computation and Language · Cross-listed from cs.AI

MuCRASP: Multimodal Chain-of-thought Reasoning aware Structured Pruning

Aritra Dutta and Somak Aditya

Vision-language models (VLMs) increasingly rely on chain-of-thought (CoT) reasoning to solve complex multimodal tasks, but their large parameter sizes make deployment expensive. Structured pruning offers a natural solution; however, existing methods fail to preserve CoT reasoning accuracy in VLMs. We identify two key reasons: (1) CoT consistency depends on sparse transition points (pivot tokens) in the generation trajectory, while existing pruning methods are CoT-agnostic; and (2) pruning methods designed for unimodal LLMs do not account for activation-distribution differences across visual and textual modalities. Motivated by these observations, we propose MuCRASP, a structured pruning framework that targets reasoning-critical components while preserving cross-modal alignment and accounting for layer-wise sensitivity under a global parameter budget. Experiments on four VLMs across three reasoning benchmarks show that MuCRASP consistently preserves reasoning quality under increasing compression. At 30% pruning on Qwen2.5-VL-7B, MuCRASP achieves an LLM-as-a-Judge score of 8.87 versus 7.32 for the strongest baseline on physical reasoning tasks. Furthermore, MuCRASP maintains high reasoning consistency up to 50% pruning, significantly outperforming prior pruning approaches while exhibiting lower perplexity degradation.
Read the full abstract

Computation and Language · Cross-listed from cs.CR

CTI2TTP: Evidence-Grounded TTP Extraction from Cyber Threat Intelligence via Diverge-then-Converge Verification

Yutong Cheng, Changze Li, Raihan Sultan Pasha Basuki et al. (+3 more)

Extracting MITRE ATT&CK techniques from cyber threat intelligence (CTI) reports is an open-set, multi-label problem requiring both high recall (not missing techniques) and high precision (not hallucinating unsupported ones). Existing rule-based, supervised, and LLM-based methods struggle to achieve both: rule-based and supervised approaches lack generalizability across diverse attack descriptions, while LLM-based approaches that couple candidate generation and validation within a single inference step suffer from limited recall and precision simultaneously. We propose CTI2TTP, which addresses this challenge through a diverge-then-converge design inspired by how human analysts work: first extracting broadly, then verifying rigorously. Reports are first decomposed into atomic behaviors, each anchored by a deterministic span localization stage to a specific evidence window in the source text. Candidate techniques are then proposed broadly for each behavior, and a convergent verification stage retains only candidates supported by both the localized evidence and the authoritative MITRE definition. We contribute two evaluation resources, a cleaned TRAM benchmark (TRAM-Clean) and a new annotated dataset (CTI2TTP-Bench), to address known annotation noise in existing benchmarks and elevate the task to document-level TTP extraction. On TRAM-Clean and CTI2TTP-Bench, CTI2TTP achieves 76.48% and 87.39% macro-F1 respectively, outperforming chain-of-thought prompting with the same backbone by 63.5% and 29.4%.
Read the full abstract

All authors Yutong Cheng, Changze Li, Raihan Sultan Pasha Basuki, Qian Cui, Wei Ding, Peng Gao

Computation and Language · Cross-listed from cs.RO

When Search Becomes Memory: Accelerating Robot Design Discovery with Self-Evolving Skills

Yunfei Wang, Xiaohao Xu, Yang Li et al. (+1 more)

Large language models (LLMs) are increasingly used as proposal generators for evolutionary robot design, yet most loops remain memoryless: simulator results shape the next population but are not preserved as reusable design knowledge. We present Auto-Robotist, a self-evolving LLM agent that distills morphology-search traces into an explicit natural-language skill library. Each skill stores a structural archetype, evidence-grounded positive and negative rules, and the evaluated designs that support them, making design memory inspectable rather than implicit in a population. During search, the agent retrieves skills to condition LLM edits of elite bodies while retaining a Genetic Algorithm (GA) mutation path for exploration; after evaluation, it updates the library through Add, Diagnose, and Merge. Across seven EvoGym tasks spanning locomotion, traversal, and object interaction, Auto-Robotist improves cold-start 5x5 search and transfers learned skills to 10x10 design spaces, where reference-conditioned transfer outperforms GA on every task. These results suggest that LLM agents can convert expensive physical evaluations into reusable, auditable design principles. Our code is publicly available at https://github.com/wangyf9/Auto-Robotist .
Read the full abstract

All authors Yunfei Wang, Xiaohao Xu, Yang Li, Xiaonan Huang

Computation and Language

Clarify, Abstain or Answer? Strategising in Conversation with Belief-Augmented Generation

Joris Baan, Wilker Aziz, Barbara Plank et al. (+1 more)

Large language models (LLMs) define a distribution over text, which can be viewed as a probabilistic representation of uncertainty: sampling K responses yields a belief state - responses a model deems plausible. Existing work exploits this representation for narrow tasks like either decoding or selective prediction, and often requires manual interventions, not controlling generation directly. We propose Belief-Augmented Generation (BAG): grounding LLMs in their own belief state via the prompt and letting them reason over these K samples to decide on and execute a conversational strategy: clarify, abstain, or answer. In a multi-turn ambiguous question answering (QA) setting, we find that LLMs by default rarely clarify or abstain, ignoring uncertainty about the input (aleatoric) or facts (epistemic). BAG improves QA accuracy across six models and yields strategy decisions more faithful to their belief state than prompt-only baselines. Disentangling when to clarify from when to abstain, however, remains challenging.
Read the full abstract

All authors Joris Baan, Wilker Aziz, Barbara Plank, Raquel Fernández

Computation and Language

Fine-Tuning Over Architectural Complexity: Broad-Coverage PII Detection on PIIBench with DeBERTa

Pritesh Jha

Personally identifiable information (PII) detection systems are frequently trained within narrow source or domain boundaries, limiting coverage when deployed on heterogeneous text. We study model fine-tuning on a corrected multi-source PIIBench preparation spanning 82 retained entity types across ten source datasets. We evaluate three DeBERTa-based approaches: direct token classification fine-tuning, a source-conditioned hierarchical model (SC+H), and a three-phase curriculum extension (SC+H+Curr). Against eight published comparator systems on a reproducible 5,000-record held-out subset (test_5k), direct fine-tuned DeBERTa achieves F1 0.6476, while SC+H and the curriculum variant achieve 0.5899 and 0.2772 respectively; the strongest published comparator reaches only 0.1723. Because validation initially favoured SC+H, we perform a final streamed evaluation on the complete 100,002-record held-out split. Direct fine-tuning remains superior, achieving F1 0.6455 versus 0.5894 for SC+H. Entity-level analysis shows that direct fine tuning wins 54 of 82 fine entity types and all ten coarse groups by support-weighted entity F1, while SC+H retains localised advantages on 28 types. The results indicate that diverse task-specific training data and a simple weighted cross-entropy objective contribute more to broad-coverage PII detection than the tested architectural and curriculum complexity.
Read the full abstract

Computation and Language

Adaptive Graph Refinement and Label Propagation with LLMs for Cost-Effective Entity Resolution

Hongtao Wang, Renchi Yang, Haoran Zheng et al. (+1 more)

Dirty entity resolution (ER), which identifies records referring to the same real-world entity from a single, messy dataset, is a fundamental task in data management and mining. However, the dominant blocking-matching-clustering paradigm for ER suffers from critical flaws. Its cascaded, decoupled workflow essentially produces a static, sparse graph plagued by missing edges (due to blocking failures) and noisy links (due to matching errors), causing error propagation and yielding suboptimal clusters, particularly when rigid transitivity is imposed in the clustering. We contend that matching and clustering are fundamentally synergistic, both optimizing for the construction of an ideal entity graph. Building upon this insight, we propose Alper, a unified framework that integrates these steps into an iterative probabilistic label propagation process over a global, evolving graph. Unlike disjoint blocking, Alper refines the graph structure and labels dynamically by adaptively integrating "weak but cheap" signals from graph propagation with "strong but expensive" LLM-based pairwise queries. For higher cost-effectiveness, we formulate the signal selection as a constrained optimization problem maximizing cumulative marginal gain under a query budget, solved via our greedy algorithm with provable theoretical guarantees. Our extensive experiments over eight benchmark datasets demonstrate that Alper is consistently superior to state-of-the-art cascaded pipelines.
Read the full abstract

All authors Hongtao Wang, Renchi Yang, Haoran Zheng, Xiangyu Ke

Computation and Language · Cross-listed from cs.CR

SAMark: A Self-Anchored Text Watermarking with Paragraph-Level Paraphrase Robustness

Jiahao Huo, Wenjie Qu, Yibo Yan et al. (+5 more)

Semantic-level watermarking (SWM) improves robustness against text modifications by treating sentences as the basic unit. However, robustness to paragraph-level paraphrasing remains difficult because such attacks globally disrupt watermark signals by changing sentence order. In this work, we propose SAMark, a self-anchored watermarking framework that removes the dependency on sentence order by establishing a step-independent green region in semantic space. To improve detectability, we introduce a multi-channel hyperbolic scoring mechanism that amplifies watermark signals while suppressing noise from weakly aligned candidates. We further propose a diversity-aware filtering strategy that combines hard filtering with soft regularization, extending beyond simple n-gram repetition filters to address semantic redundancy. Experimental results show that SAMark achieves up to 90.2% TP@FP1% under typical paragraph-level paraphrasing attacks, outperforming the strongest prior baseline by more than 30% on average, while maintaining generation quality competitive with unwatermarked text and breaking the robustness-quality trade-off that limits prior methods. Our code will be released at [this URL](https://github.com/Z1zs/SAMark).
Read the full abstract

All authors Jiahao Huo, Wenjie Qu, Yibo Yan, Kening Zheng, Jiaheng Zhang, Xuming Hu, Philip S. Yu, Mingxun Zhou

Computation and Language

Double Triangle Annotation: A Scalable Human-in-the-Loop Framework for High-Precision Historical Document Annotation

Yi Ren

Evaluating structured-information extraction from historical documents at scale requires high-precision ground-truth annotations, yet traditional manual labeling is expensive and fully automated pipelines built on large language models are prone to hallucination. We propose Double Triangle Annotation, a two-layer human-in-the-loop framework that leverages cross-model consensus to automate the majority of annotation work while ensuring high-precision outputs. In the first layer, two architecturally independent Multimodal Large Language Models annotate each document in parallel; when they agree, the label is auto-accepted, and disagreements are routed to a human jury. A second layer cross-checks two such systems against each other, escalating residual conflicts to a domain expert. The framework rests on a single assumption – error independence between models – requires no distributional priors or task-specific calibration, and becomes more autonomous as model capability improves. On the Guides Rosenwald, a corpus of French medical directories spanning 1887-1906, the framework achieves a final Word Error Rate of 0.003. Applied at scale, model consensus auto-accepts over 85% of 13,595 fields. We release the resulting benchmark – the first structured-extraction ground truth for the Rosenwald Guides – to support future work on historical document processing.
Read the full abstract

Computation and Language · Cross-listed from stat.ML

Efficient Benchmarking Is Just Feature Selection and Multiple Regression

Sam Bowyer, Acyr Locatelli and Kris Cao

Efficient benchmarking techniques aim to lower the computational cost of evaluating LLMs by predicting full benchmark scores using only a subset of a benchmark's questions. By reframing this problem as an instance of multiple regression with feature selection, we find that existing efficient benchmarking methods can be greatly improved by simply using kernel ridge regression at the prediction stage. Additionally, using an information-theoretic feature-selection algorithm called minimum redundancy maximum relevance (mRMR), we can further improve upon these methods by selecting question subsets that will be maximally useful for prediction. Except in very data-poor settings, these approaches consistently achieve smaller prediction errors (in both MAE and RMSE), and greater ranking correlation between predicted and true scores (in both Spearman $\rho$ and Kendall $\tau$) across a range of benchmarks using both binary and continuous metrics. Furthermore, mRMR subsampling is much faster than competitor methods (which often involve fitting probabilistic models or running clustering algorithms), and is more likely to select the same questions under different random seeds or training data splits. Tutorial code can be found at https://github.com/sambowyer/mrmr_eval .
Read the full abstract

Computation and Language

StreamProfileBench: A Benchmark for Fine-Grained User Profile Inference in Real-World Streaming Scenarios

Sizhe Wang, Feiyu Duan, Juelin Wang et al. (+2 more)

Large Language Models (LLMs) have reshaped user profiling, yet current evaluations mainly focus on static data snapshots. This paradigm overlooks the reality of personalized systems, where User-Generated Content (UGC) arrives continuously and fine-grained profiles evolve rapidly. To bridge this gap, we introduce StreamProfileBench, a large-scale benchmark for fine-grained streaming user profiling. We formalize streaming user profiling as a continuous state maintenance task and curate a highly authentic dataset comprising over 120,000 UGC posts from 7,000+ real users across five diverse platforms. By leveraging the temporal correlation of user interests, we further propose a novel, annotation-free evaluation framework. Extensive experiments across 14 leading LLMs reveal that continuous profile updating remains an open challenge. Models exhibit a systemic conservative bias, over-retaining past interests while failing to recognize interest decay. Ablation experiments further validate the practical utility and necessity of the streaming paradigm. Data and code are hosted in https://github.com/WaterWang-001/StreamProfileBench.
Read the full abstract

All authors Sizhe Wang, Feiyu Duan, Juelin Wang, Liwen Zhang, Zhongyu Wei

Computation and Language

Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chains

Hui Xie, Jie Liu, Ziyue Qiao et al. (+1 more)

Explicit chain-of-thought (CoT) reasoning substantially improves the reasoning ability of large language models (LLMs), but incurs high inference cost due to lengthy autoregressive traces. Existing latent reasoning methods offer a promis- ing alternative, yet they often treat reasoning as uniformly compressible, causing precision-critical intermediate steps to be overly compressed and thereby degrading reasoning accuracy. In this work, we propose Selective Latent Thinking (SLT), a reasoning paradigm that selectively compresses redundant reasoning spans into latent representations while preserving precision-critical spans as explicit CoT within the same reasoning trajectory. Our primary instantiation uses a lightweight decoder to anticipate a short upcoming reasoning span and applies confidence- based gating to determine the longest span that can be reliably compressed; the resulting selective compression policy is further optimized with trajectory-level reinforcement learning to balance answer correctness against reasoning cost. Ex- periments across four mathematical reasoning benchmarks show that which spans are compressed is critical: with the same compressor, compressing arbitrary spans reduces the GSM-Hard accuracy of Qwen3-4B from 54.7% to 17.3%, whereas keeping numerical tokens explicit retains 51.8%. SLT reduces explicit reasoning length by 58.4% with only a 2.8-point accuracy drop compared to explicit CoT on Llama-3.2-1B. Notably, the principle extends beyond this design: instantiated within existing latent reasoning models via a learned compression gate, it improves accuracy at equal reasoning length (+2.6 points over Latent-SFT) and at shorter length (+5.6 points over Latent-GRPO with 13% fewer tokens).
Read the full abstract

All authors Hui Xie, Jie Liu, Ziyue Qiao, Joaquin Vanschore

Computation and Language

Trait-Aware Policy Optimization for Autoregressive Multi-Trait Essay Scoring

Zhengyang Wang, Sanwoo Lee, Jiaxin Wang et al. (+3 more)

Multi-trait essay scoring aims to provide fine-grained evaluation of writing quality across multiple dimensions. However, how to effectively post-train autoregressive scoring models remains underexplored. In this paper, we propose Trait-Aware Policy Optimization (TAPO), a post-training framework tailored to autoregressive multi-trait scoring. Our method decomposes rewards along both the sample and trait dimensions, combining global scoring consistency, trait-level accuracy, format validity, and inter-trait dependency preservation. In addition, we use enhanced prompts throughout training by incorporating original prompt texts and trait descriptions, providing richer semantic information for trait-specific score generation. Experiments across multiple backbone models show that our method consistently improves multi-trait scoring performance over supervised fine-tuning and scalar-reward optimization baselines, demonstrating the effectiveness and transferability of trait-aware post-training for essay scoring.
Read the full abstract

All authors Zhengyang Wang, Sanwoo Lee, Jiaxin Wang, Chenxi Miao, Weikang Li, Yunfang Wu

Computation and Language · Cross-listed from cs.CV

CMAP: Cross-Modal Adaptive Prompting for Multi-Domain Task-Incremental Learning

Sriram Mandalika

Multi-domain task-incremental learning requires a model to sequentially acquire knowledge across visually diverse domains without forgetting prior tasks, and without access to task identity at inference. Parameter-efficient methods built on frozen vision-language models have made strong progress, yet all existing approaches rely exclusively on visual features for task routing, confidence estimation, and encoder adaptation, leaving CLIP's cross-modal text embedding space entirely unexploited. We address this gap through three contributions. Text-space task routing replaces visual Gaussian matching with cosine similarity to frozen CLIP text prototypes, giving order-independent routing robust to data scarcity at zero parameter cost. Multi-prototype visual-textual confidence replaces single-Gaussian class modeling with K-means visual prototypes and cross-modal alignment scores under task-calibrated thresholds. Symmetric cross-modal gating extends per-layer Gumbel gates to the text encoder conditioned on batch image features, preserving cross-modal alignment on out-of-distribution inputs. On the MTIL benchmark spanning 11 datasets and 1201 classes, our method achieves 74.2% Transfer, 80.5% Average, and 88.7% Last under Order-I, surpassing the prior state of the art by 5.0, 3.7, and 3.0 percentage points with only 2.5M trainable parameters and no external data.
Read the full abstract

Computation and Language

PowLU: An Activation Function for Stable Pre-Training of LLMs

Peijie Jiang, Yuqi Feng, Cunyin Peng et al. (+5 more)

In contemporary large language models (LLMs), the swish-gated linear unit (SwiGLU) activation function is widely adopted to regulate the information flow and introduce non-linearity. For large positive inputs, SwiGLU approximates the quadratic function $x^2$, providing strong nonlinearity and expressive capacity. However, this property also causes numerical instability as the input or model scale increases, particularly in low-precision LLM training. The main reason is its approximate quadratic amplification, which enlarges the output range and exacerbates outliers. To address this issue, we propose a stable activation function, Power Linear Unit (PowLU), for large-scale LLM pre-training. Specifically, PowLU employs a rational power function to achieve adaptive nonlinearity, thereby improving representation ability and enabling stable training in spike regions. Moreover, we provide theoretical justification for several key properties of PowLU. Scaling law experiments confirm that the performance is consistent across model sizes, and further experimental results with the Ling architecture (7.9B and 124B total parameters) demonstrate that PowLU achieves competitive results against SwiGLU and SwiGLU-Clip in large-scale training of LLMs. In addition, the experimental results also show that PowLU effectively improves the scalability of the large-scale training of LLMs.
Read the full abstract

All authors Peijie Jiang, Yuqi Feng, Cunyin Peng, Qian Zhao, Jia Liu, KunLong Chen, Zhiqiang Zhang, Jun Zhou

Computation and Language · Cross-listed from cs.DC

Neural Router: Semantic Content Matching for Agentic AI

Lauri Lovén, Abhishek Kumar, Alexander Engelhardt et al. (+5 more)

Large language models (LLMs) can serve as the semantic-matching engine of a content-based publish/subscribe broker for agentic AI across the edge-cloud computing continuum, bridging the vocabulary and modality gaps that defeat keyword and embedding filters. Framed as offline multi-label retrieval over three public datasets spanning social-media, legal, and smart-home sensor domains (six LLMs, seven baselines), our central contribution is a two-crossover cost-accuracy characterisation: an analytical context-window crossover below which a CoverAndMerge compression pipeline reduces LLM invocations, and an empirical discrimination-capacity crossover above which matching accuracy collapses independently of context budget, by a model-dependent factor of parameter count and training generation. Two findings carry practical weight: above the discrimination crossover, compression cannot recover accuracy and only frontier-scale models clear large subscription sets; and there backend choice dominates configuration choice, so model selection, not pipeline tuning, is the primary operator lever. We accompany this with three composable algorithms and a per-cluster Quality-of-Experience framework for autonomic LLM-tier selection.
Read the full abstract

All authors Lauri Lovén, Abhishek Kumar, Alexander Engelhardt, Alaa Saleh, Roberto Morabito, Xiaoli Liu, Naser Hossein Motlagh, Sasu Tarkoma

Computation and Language

From Facts to Insights: A Persona-Driven Dual Memory Framework and Dataset for Role-Playing Agents

Rongsheng Zhang, Ruofan Hu, Weijie Chen et al. (+7 more)

While role-playing agents excel in short-term interactions, long-term conversations overwhelm context windows, motivating external memory frameworks. Current systems typically rely on persona-agnostic summarization, which records facts without persona-specific interpretation, yielding generic responses that compromise persona fidelity. To bridge this gap, we introduce RoleMemo, a dataset featuring four reasoning tasks where the factual fragments must be interpreted through the persona to reach the correct answer. Evaluation on RoleMemo exposes critical limitations of persona-agnostic frameworks. We thus propose DualMem, which decouples memory into two streams: factual cognition and persona-conditioned insight. Trained through Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), our framework with a 4B-parameter model outperforms zero-shot persona-agnostic frameworks powered by DeepSeek-V3.2 for sustained persona fidelity. Our resources are available at https://github.com/role2026/rolememo.
Read the full abstract

All authors Rongsheng Zhang, Ruofan Hu, Weijie Chen, Jiji Tang, Junnan Ren, Wanying Wu, Xunuoyan Chen, Tangjie Lv, Tao Jin, Zhou Zhao

Computation and Language

Testing the Deliteralization Hypothesis in Human and Machine Translation

Malik Marmonier, Rachel Bawden and Benoît Sagot

The recent shift from dedicated NMT systems to general-purpose LLMs has reshaped machine translation, with LLMs reported to produce more fluent, less literal output than their predecessors. We test whether this shift extends to the deliteralization hypothesis, the long-standing claim from translation studies that translations become progressively less literal as they are drafted and revised. Using the WMT24++ dataset, we compare the literality of human translations and post-editions to that of two NMT systems and six LLMs across 54 language pairs and three tasks: direct translation, iterative self-revision, and post-editing of human drafts. Literality is measured via a validated Synthetic Literality Index built from six heuristics. We find that (i) human translations remain significantly less literal than those of all tested MT systems, though recent LLMs narrow the gap; (ii) when prompted to iteratively revise their own output, LLMs deliteralize monotonically, providing the first evidence that the hypothesis applies natively to LLM generation; and (iii) as post-editors, LLMs invert the revision triggers of human post-editors, tolerating literal drafts and targeting idiomatic human formulations for revision.
Read the full abstract

Computation and Language

Simulating Human Memory with Language Models

Qihan Wang, Nicholas Tomlin, Michael Hu et al. (+2 more)

Language models are increasingly being deployed as user simulators, but their memory is far more reliable than that of real users. To measure this gap, we run a series of classic memory experiments from psychology on both humans and language models. Across tasks, we find that out-of-the-box language models exhibit better memory than humans, even when prompted to imitate human behavior. We then show that better prompting strategies and the use of a compactor can cause language models to forget content in a more human-like way. Using these methods, we show preliminary evidence that language models with human-like memory constraints can function as more effective user simulators in a downstream education task. Finally, we release human reference data and benchmarks to support future work on simulating human memory with language models.
Read the full abstract

All authors Qihan Wang, Nicholas Tomlin, Michael Hu, Brian Dillon, Tal Linzen

Computation and Language

Llamion Technical Report

Kisu Yang, Yoonna Jang, Hyeonseok Moon et al. (+6 more)

We release Llamion, a family of 14B-parameter open-weight language models obtained by transforming Orion-14B into the standardized Llama-family architecture. The transformation is performed by Efficient Knowledge Preservation for Transformation (KEPT), a recipe that combines (i) Normal Parameter Mapping (NPM) for unchanged modules, (ii) Optimized Parameter Mapping (OPM), a training-free LayerNorm-to-RMSNorm initialization we prove optimal under the near-zero-mean activation regime induced by weight decay, and (iii) Cross-architecture Knowledge Distillation (XKD), an equal-size frozen-teacher distillation that aligns the converted model's outputs with the source model's on any reasonable input distribution. Llamion recovers Orion's behaviour on H6, MT-Bench, and KoMMLU with only ~123M tokens on a single A100 in four days; Llamion-Base reaches 66.87% on KoMMLU, exceeding the next-best entry of the Open Ko LLM Leaderboard by >7.0 absolute points at submission time. Capabilities entirely absent from the transfer corpus (Python programming and 200K-token context handling) survive the architectural transition intact. We release three checkpoints (Base, Chat, LongChat) that load with trust_remote_code=False in the Hugging Face Transformers library.
Read the full abstract

All authors Kisu Yang, Yoonna Jang, Hyeonseok Moon, Hwanseok Jang, Taewoo Lee, Hyungjin Lee, Jeseung Lee, Juhyoung Park, Heuiseok Lim

Computation and Language

AutoSG: LLM-Driven Solver Generation Solely from Task Prompts for Expensive Optimization

Haoran Gu, Handing Wang, Yi Mei et al. (+1 more)

Expensive optimization tasks are ubiquitous in real-world applications, demanding highly specialized solvers. While LLM-driven automated solver generation shows promise, current paradigms face three critical issues when tackling expensive optimization: factual hallucinations due to deficient domain knowledge, the frequent dismantling of previously established locally optimal structures during refinement, and the prohibitive evaluation costs alongside restricted generalization caused by executing on training instances. To address these issues, we introduce AutoSG, a fully automated workflow directly translating natural language prompts into executable customized solvers. AutoSG features three core innovations: a retrieval-augmented solver generation module strictly grounding code in verified literature; a one-step self-refinement operator introducing task-specific improvements while preserving critical structural components; and an instance-free Elo-based LLM-as-a-Judge evaluation mechanism rapidly establishing global rankings. Extensive evaluations across diverse expensive optimization tasks confirm AutoSG significantly outperforms human-designed state-of-the-art frameworks and existing LLM-generated solvers.
Read the full abstract

All authors Haoran Gu, Handing Wang, Yi Mei, Mengjie Zhang

Computation and Language

A Two-Phase Stability Study of LLM Judges and Bar Council Examiners on Thai Bar-Exam Free-Form Essays

Pawitsapak Akarajaradwong, Wuttikrai Lertprasertphakorn, Chompakorn Chaksangchaichot et al. (+1 more)

Free-form legal essay evaluation in NLP treats expert inter-rater stability as a single ceiling number, and treats LLM-judge agreement with that ceiling as evidence of judge stability. We test both assumptions on the Thai bar examination through an identical-inputs protocol: three Bar Council-trained examiners (A, B, C) and a 26-LLM judge panel score the same 15 cross-graded answers from the same four inputs (question, official Bar Council grading regulation, gold answer, candidate answer). The headline finding is asymmetric. On 10 of 15 cells where the rubric prescribes both axes, all 29 raters converge in a tight band: panel agreement is universal. On the remaining 5 cells where the rubric does not prescribe how to grade a correct final answer that omits a decisive statutory citation, the human panel splits between two coherent readings (B/C majority at the upper rubric band, score 6-8; A minority at the lower band, score 1-2). The LLM judge population does not split symmetrically: 22 of 26 LLMs score in or near B/C's contested band, 3 sit in the regulation-silent middle gap, and only 1 (GPT-5.4 Nano) approaches A's band without consistently scoring within it. Zero LLMs in our 26-judge panel reproduce the minority human reading on the contested cells. The B/C-direction cluster spans every model size, vendor, and price tier we tested. An instrumented three-LLM anchor sub-panel (Claude 4.6 Opus, Gemini 3.1 Pro, GPT-5.4 Pro) carries determinism probes, input ablations, and bootstrap CIs, and reaches anchor panel $\alpha = 0.77$ on the 15 cells against human-panel $\alpha = 0.36$. The high LLM-panel $\alpha$ reflects systematic convergence on the majority reading rather than balanced reproduction of both readings; a benchmark that selects its LLM judge by maximising agreement with a human reference panel will inherit this asymmetry by construction.
Read the full abstract

All authors Pawitsapak Akarajaradwong, Wuttikrai Lertprasertphakorn, Chompakorn Chaksangchaichot, Sarana Nutanong

Computation and Language

Iterate Until Retrieved: Factual Nugget Optimization for Discoverable Continual Corrections in Agentic RAG

Moshe Hazoom, Gal Patel, Alon Talmor et al. (+1 more)

Agentic retrieval-augmented generation (RAG) systems in business-to-business settings may often receive free-form response feedback. Rather than generic feedback signals such as style, preference, or overall response quality, we focus on actionable factual corrections. We identify these instances and convert them into compact knowledge-base entries, which we call factual nuggets. We introduce Iterative Nugget Optimization (INO), an index-time optimization method that uses the production agentic RAG as a test harness: it creates an initial nugget, probes it with the triggering query and paraphrases, reflects over failed retrieval and answer traces, and revises the nugget until it is discoverable. We evaluate INO with two production B2B knowledge-assistance agents across multiple companies that use our system: a product support agent that answers questions over company-specific knowledge bases, and a support ticket agent that assists support engineers. INO consistently improves results over baselines in terms of discoverability and usage of factual corrections, in automated and human evaluations. A controlled experiment in production shows that the nuggets improve user satisfaction with responses.
Read the full abstract

All authors Moshe Hazoom, Gal Patel, Alon Talmor, Tom Hope

Computation and Language

Reinforcement Learning from Denoising Feedback

Qi He, Huan Chen, Ya Guo et al. (+3 more)

Policy loss estimation remains a fundamental and long-standing challenge in reinforcement learning (RL) for diffusion language models (DLMs). We introduce Reinforcement Learning from Denoising Feedback (RLDF), a novel training paradigm that leverages feedback obtained from rollout and training processes to facilitate accurate and efficient policy loss estimation. To balance the trade-off between computational efficiency and estimation effectiveness, RLDF optimizes the model toward the clipped clean state from intermediate noisy states, combined with weighted timestep sampling over denoising timesteps. Extensive experiments demonstrate that RLDF achieves consistent and substantial improvements in both performance and generalizability across two representative DLM architectures, LLaDA and Dream, on multiple reasoning benchmarks. Our work lays a principled foundation for scalable reinforcement learning in diffusion language models. We build Drift, a training framework for DLMs, available at https://github.com/ant-research/Drift.
Read the full abstract

All authors Qi He, Huan Chen, Ya Guo, Huijia Zhu, Yi R. Fung, Baojian Zhou