Computation and Language
ImCoref-CeS: An Improved Lightweight Pipeline for Coreference Resolution with LLM-based Checker-Splitter Refinement
Kangyang Luo, Yuzhuo Bai, Shuzheng Si et al. (+9 more)
Coreference Resolution (CR) is a critical task in Natural
Language Processing (NLP). Current research faces a key dilemma: whether to further explore the potential of supervised neural methods based on small
language models, whose detect-then-cluster pipeline still delivers top performance, or embrace the powerful capabilities of Large
Language Models (LLMs). However, effectively combining their strengths remains underexplored. To this end, we propose ImCoref-CeS, a novel framework that integrates an enhanced supervised model with LLM-based
reasoning. First, we present an improved CR method (ImCoref) to push the performance boundaries of the supervised neural method by introducing a lightweight bridging module to enhance long-text encoding capability, devising a biaffine scorer to comprehensively capture positional information, and invoking a hybrid mention regularization to improve training efficiency. Importantly, we employ an LLM acting as a multi-role Checker-Splitter
agent to validate candidate mentions (filtering out invalid ones) and coreference results (splitting erroneous clusters) predicted by ImCoref. Extensive experiments demonstrate the effectiveness of ImCoref-CeS, which achieves superior performance compared to existing state-of-the-art (SOTA) methods.
Read the full abstract ▾
All authors Kangyang Luo, Yuzhuo Bai, Shuzheng Si, Cheng Gao, Zhitong Wang, Yingli Shen, Wenhao Li, Zhu Liu, Yufeng Han, Jiayi Wu, Cunliang Kong, Maosong Sun
Computation and Language
Text2Token: Unsupervised Text Representation Learning with Token Target Prediction
Ruize An, Richong Zhang, Zhijie Nie et al. (+3 more)
Unsupervised text representation learning (TRL) is a fundamental task in natural
language processing, which is beneficial for improving search and recommendations with the web's unlabeled texts. A recent empirical study finds that the high-quality representation aligns with the key
token of the input text, uncovering the potential connection between representation space and vocabulary space. Inspired by the findings, we revisit the generative tasks and develop an unsupervised generative framework for TRL, Text2
Token. The framework is based on the
token target prediction task, utilizing carefully constructed target
token distribution as supervisory signals. To construct the high-quality target
token distribution, we analyze the
token-
alignment properties with advanced embedders and identify two essential categories of key
tokens: (1) the meaningful
tokens in the text and (2) semantically derived
tokens beyond the text. Based on these insights, we propose two methods – data-driven and model-derived – to construct synthetic
token targets from data or the LLM backbone. Experiments on the MTEB v2
benchmark demonstrate that Text2
Token achieves performance competitive with the state-of-the-art embedder with unsupervised contrastive learning, LLM2Vec. Our analysis further shows that vocabulary and representation spaces optimize together and toward the optimum solution during training, providing new ideas and insights for future work.
Read the full abstract ▾
All authors Ruize An, Richong Zhang, Zhijie Nie, Zhanyu Wu, Yanzhao Zhang, Dingkun Long
Computation and Language
You only need 4 extra tokens: Synergistic Test-time Adaptation for LLMs
Yijie Xu, Huizai Yao, Zhiyu Guo et al. (+5 more)
Large
language models (LLMs) are increasingly deployed in specialized domains such as finance, medicine, and agriculture, where they face significant distribution shifts from their training data. Domain-specific fine-tuning can mitigate this challenge but relies on high-quality labeled data that is expensive and slow to collect in expertise-limited settings. We study label-free test-time adaptation for
language models and present SyTTA, an inference-time framework that adapts models on-the-fly without additional supervision. SyTTA couples two complementary uncertainty signals that arise under distribution shift: input-side perplexity, indicating mismatch with domain-specific terminology and patterns, and output-side predictive
entropy, indicating diffuse and unstable
token probabilities during generation. Across diverse model architectures and domain-specific
benchmarks, SyTTA delivers consistent gains. Notably, on agricultural question answering, SyTTA improves Rouge-LSum by over 120% on Qwen-2.5-7B with only 4 extra
tokens per query. These results show that effective test-time adaptation for
language models is achievable without labeled examples, supporting deployment in label-scarce domains. The code will be made available upon acceptance.
Read the full abstract ▾
All authors Yijie Xu, Huizai Yao, Zhiyu Guo, Pengteng Li, Aiwei Liu, Xuming Hu, Weiyu Guo, Hui Xiong
Computation and Language
Weed Out, Then Harvest: Dual Low-Rank Adaptation is an Effective Noisy Label Detector for Noise-Robust Learning
Bo Yuan, Yulin Chen and Yin Zhang
Parameter-efficient fine-tuning (PEFT) large
language models (LLMs) have shown impressive performance in various downstream tasks. However, in many real-world scenarios, the collected training data inevitably contains noisy labels. To learn from noisy labels, most solutions select samples with small losses for model training. However, the selected samples, in turn, impact the loss computation in the next iteration. An inaccurate initial selection can create a vicious cycle, leading to suboptimal performance. To break this cycle, we propose Delora, a novel framework that decouples the sample selection from model training. For sample selection, Delora establishes a noisy label detector by introducing clean and noisy LoRA. Benefiting from the
memory effect, the clean LoRA is encouraged to memorize clean data, while the noisy LoRA is constrained to memorize mislabeled data, which serves as a learnable threshold for selecting clean and noisy samples. For model training, Delora can use carefully selected samples to fine-tune
language models seamlessly. Experimental results on synthetic and real-world noisy
datasets demonstrate the effectiveness of Delora in noisy label detection and text classification.
Read the full abstract ▾
Computation and Language · Cross-listed from cs.LG
RLFR: Extending Reinforcement Learning for LLMs with Flow Environment
Jinghao Zhang, Naishan Zheng, Ruilin Li et al. (+4 more)
Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as a promising framework for improving
reasoning abilities in Large
Language Models (LLMs). However, policy optimized with binary verification prone to overlook potential valuable exploration in
reasoning trajectory. In view of heavy annotation cost of golden Process Reward Models (PRMs), recent works attempt using auxiliary signals for reward shaping of process
tokens, involving
entropy and likelihood collected from logit space. In this work, we offer a novel perspective on shaping RLVR with flow rewards derived from latent space, and propose RLFR, where the flow fields of model latents are constructed from either off-policy high-quality data and on-policy rejection sampling data, and the velocity deviations of policy latents within it are quantified to serve as a reward signal. RLFR first demonstrates that a well-established flow field can be a sound environment for reward signal collection, highlighting the expressive latent space is much underexplored. Moreover, RLFR is able to compress any off-policy expert data as reference for constituting reward signals, and we show that the efficient context dependence compressed within the hidden states are utilized, rather than individual
token-level denotation for context comprehending. Experiments on both
language and
multimodal reasoning benchmarks demonstrate the reliability of flow rewards, and suggesting a promising paradigm for reward shaping with auxiliary signals.
Read the full abstract ▾
All authors Jinghao Zhang, Naishan Zheng, Ruilin Li, Dongzhou Cheng, Zheming Liang, Feng Zhao, Jiaqi Wang
Computation and Language
Auditing medical multi-agent AI reveals risks of false consensus
Yinghao Zhu, Lei Gu, Zixiang Wang et al. (+11 more)
Large
language models are increasingly being assembled into medical multi-
agent systems that emulate multidisciplinary consultation through specialist roles, peer review and consensus formation. In clinical decision support, however, apparent consensus is not enough. Clinicians also need to know whether
agents checked the evidence, addressed disagreement and kept uncertainty visible. Current evaluations largely score final accuracy, leaving the safety of the collaborative process untested. Here we introduce MedAgentAudit, a clinically grounded workflow audit framework for diagnosing and quantifying collaborative failure modes in medical multi-
agent systems. From 3,600 execution logs, we derive an expert-validated taxonomy of ten recurrent failures spanning task comprehension, collaborative discussion, and synthesis and decision-making. We then deploy an expert-validated automated auditor as non-interventional probes across 14,400 cases, covering six multi-
agent architectures, six medical text and
vision datasets, and four large
language model settings per modality. Across systems, collaboration yields uneven accuracy gains and frequent process failures. Unsupported observations affect 16.63% of cases and propagate downstream. In discussion,
agents repeat initial views in 98.42% of cases rather than re-examining evidence, and fail to activate specialist
reasoning in 42.73%. During synthesis, final answers often substitute authority or majority count for evidence checking, showing authority bias in 28.76% (rising from 35.30% to 68.75% across rounds), self-contradiction in 18.53%, contradiction neglect in 5.48% and minority suppression in 5.11%. MedAgentAudit reframes medical AI evaluation from output scoring to process-level safety and accountability, providing a practical foundation for transparent, auditable and clinician-supervised agentic systems in medicine.
Read the full abstract ▾
All authors Yinghao Zhu, Lei Gu, Zixiang Wang, Haoran Sang, Dehao Sui, Wen Tang, Lan Mi, Yasha Wang, Junyi Gao, Liang Yao, Tianfan Fu, Ewen Harrison, Lequan Yu, Liantao Ma
Computation and Language
A Survey of Inductive Reasoning for Large Language Models
Kedi Chen, Dezhao Ruan, Yuhao Dan et al. (+12 more)
Reasoning is an important task for large
language models (LLMs). Among all the
reasoning paradigms, inductive
reasoning is one of the fundamental types, which is characterized by its particular-to-general thinking process and the non-uniqueness of its answers. The inductive mode is crucial for knowledge generalization and aligns better with human cognition, so it is a fundamental mode of learning, hence attracting increasing interest. Despite the importance of inductive
reasoning, there is no systematic summary of it. Therefore, this paper presents the first comprehensive survey of inductive
reasoning for LLMs. First, methods for improving inductive
reasoning are categorized into three main areas: post-training, test-time scaling, and data augmentation. Then, current
benchmarks of inductive
reasoning are summarized, and a unified sandbox-based evaluation approach with the observation coverage metric is derived. Finally, we offer some analyses regarding the source of inductive ability and how simple model architectures and data help with inductive tasks, providing a solid foundation for future research.
Read the full abstract ▾
All authors Kedi Chen, Dezhao Ruan, Yuhao Dan, Yaoting Wang, Siyu Yan, Xuecheng Wu, Yinqi Zhang, Qin Chen, Jie Zhou, Liang He, Biqing Qi, Linyang Li, Qipeng Guo, Xiaoming Shi, Wei Zhang
Computation and Language
Large Language Model Sourcing: A Survey
Liang Pang, Jia Gu, Sunhao Dai et al. (+7 more)
Due to the black-box nature of large
language models (LLMs) and the realism of their generated content, issues such as
hallucinations, bias, unfairness, and copyright infringement have become significant. In this context, sourcing information from multiple perspectives is essential. This survey presents a systematic investigation organized around four interrelated dimensions: Model Sourcing, Model Structure Sourcing, Training Data Sourcing, and External Data Sourcing. Moreover, a unified dual-paradigm taxonomy is proposed that classifies existing sourcing methods into prior-based (proactive traceability
embedding) and posterior-based (retrospective inference) approaches. Traceability across these dimensions enhances the transparency, accountability, and trustworthiness of LLMs deployment in real-world applications.
Read the full abstract ▾
All authors Liang Pang, Jia Gu, Sunhao Dai, Zihao Wei, Zenghao Duan, Kangxi Wu, Zhiyi Yin, Jun Xu, Huawei Shen, Xueqi Cheng
Computation and Language
BabyBabelLM: A Multilingual Benchmark of Developmentally Plausible Training Data
Jaap Jumelet, Abdellah Fourtassi, Akari Haga et al. (+23 more)
We present BabyBabelLM, a
multilingual collection of
datasets modeling the
language a person observes from birth until they acquire a native
language. We curate developmentally plausible pretraining data aiming to cover the equivalent of 100M English words of content in each of 45
languages. We compile evaluation suites and train baseline models in each
language. BabyBabelLM aims to facilitate
multilingual pretraining and cognitive modeling.
Read the full abstract ▾
All authors Jaap Jumelet, Abdellah Fourtassi, Akari Haga, Bastian Bunzeck, Bhargav Shandilya, Diana Galvan-Sosa, Faiz Ghifari Haznitrama, Francesca Padovani, Francois Meyer, Hai Hu, Julen Etxaniz, Laurent Prévot, Linyang He, María Grandury, Mila Marcheva, Negar Foroutan, Nikitas Theodoropoulos, Pouya Sadeghi, Siyuan Song, Suchir Salhan, Susana Zhou, Yurii Paniv, Ziyin Zhang, Arianna Bisazza, Alex Warstadt, Leshem Choshen
Computation and Language
BILLY: Steering Large Language Models via Merging Persona Vectors for Creative Generation
Tsung-Min Pai, Jui-I Wang, Li-Chun Lu et al. (+3 more)
Multi-LLM systems enhance the creativity of large
language models by simulating human collective intelligence but suffer from significant drawbacks, such as high computational costs and inference latency. To address these limitations, we propose BILLY (BlendIng persona vectors for Large
Language model creativitY), a training-free framework that captures the benefits of multi-LLM collaboration, i.e. inducing diverse perspectives and specialized expertise, within a single model. BILLY operates by extracting and blending multiple distinct persona vectors directly in the model's activation space. We steer the model's generation process with this merged vector while inference, enabling multi-perspective output without explicit multi-LLM communication. Our experiments across creativity-oriented
benchmarks demonstrate that BILLY surpasses single model prompting and traditional multi-LLM approaches, while substantially reducing inference time and computational costs. Our analyses further reveal that distinct persona vectors can be blended to achieve both effective control over complementary aspects of generation and greater interpretability.
Read the full abstract ▾
All authors Tsung-Min Pai, Jui-I Wang, Li-Chun Lu, Shao-Hua Sun, Hung-Yi Lee, Kai-Wei Chang
Computation and Language
Debiasing LLMs by Masking Unfairness-Driving Attention Heads
Tingxu Han, Wei Song, Ziqi Ding et al. (+6 more)
Large
language models (LLMs) increasingly mediate decisions in domains where unfair treatment of demographic groups is unacceptable. Existing work probes when biased outputs appear, but gives little insight into the mechanisms that generate them, leaving existing mitigations largely fragile. In this paper, we conduct a systematic investigation LLM unfairness and propose DiffHeads, a lightweight debiasing framework for LLMs. We first compare Direct-Answer (DA) prompting to Chain-of-Thought (CoT) prompting across eight representative open- and closed-source LLMs. DA will trigger the nature bias part of LLM and improve measured unfairness by 534.5%-391.9% in both one-turn and two-turn dialogues. Next, we define a
token-to-head contribution score that traces each
token's influence back to individual
attention heads. This reveals a small cluster of bias heads that activate under DA but stay largely dormant with CoT, providing the first causal link between prompting strategy and bias emergence. Finally, building on this insight, we propose DiffHeads that identifies bias heads through differential activation analysis between DA and CoT, and selectively masks only those heads. DiffHeads reduces unfairness by 49.4%, and 40.3% under DA and CoT, respectively, without harming model utility.
Read the full abstract ▾
All authors Tingxu Han, Wei Song, Ziqi Ding, Ziming Li, Chunrong Fang, Yuekang Li, Dongfang Liu, Zhenyu Chen, Zhenting Wang
Computation and Language
Hybrid OCR-LLM Framework for Enterprise-Scale Document Information Extraction Under Copy-heavy Task
Zilong Wang and Xiaoyu Shen
Information extraction from copy-heavy documents, characterized by massive volumes of structurally similar content, represents a critical yet understudied challenge in enterprise document processing. We present a systematic framework that strategically combines OCR engines with Large
Language Models (LLMs) to optimize the accuracy-efficiency trade-off inherent in repetitive document extraction tasks. Unlike existing approaches that pursue universal solutions, our method exploits document-specific characteristics through intelligent strategy selection. We implement and evaluate 25 configurations across three extraction paradigms (direct, replacement, and table-based) on identity documents spanning four formats (PNG, DOCX, XLSX, PDF). Through table-based extraction methods, our adaptive framework delivers outstanding results: F1=1.0 accuracy with 0.97s latency for structured documents, and F1=0.997 accuracy with 0.6 s for challenging image inputs when integrated with PaddleOCR, all while maintaining sub-second processing speeds. The 54 times performance improvement compared with
multimodal methods over naive approaches, coupled with format-aware routing, enables processing of heterogeneous document streams at production scale. Beyond the specific application to identity extraction, this work establishes a general principle: the repetitive nature of copy-heavy tasks can be transformed from a computational burden into an optimization opportunity through structure-aware method selection.
Read the full abstract ▾
Computation and Language
LinearRAG: Linear Graph Retrieval Augmented Generation on Large-scale Corpora
Luyao Zhuang, Shengyuan Chen, Yilin Xiao et al. (+5 more)
Retrieval-Augmented Generation (RAG) is widely used to mitigate
hallucinations of Large
Language Models (LLMs) by leveraging external knowledge. While effective for simple queries, traditional RAG systems struggle with large-scale, unstructured corpora where information is fragmented. Recent advances incorporate knowledge graphs to capture relational structures, enabling more comprehensive
retrieval for complex, multi-hop
reasoning tasks. However, existing graph-based RAG (GraphRAG) methods rely on unstable and costly relation extraction for graph construction, often producing noisy graphs with incorrect or inconsistent relations that degrade
retrieval quality. In this paper, we revisit the pipeline of existing GraphRAG systems and propose LinearRAG (Linear Graph-based
Retrieval-Augmented Generation), an efficient framework that enables reliable graph construction and precise passage
retrieval. Specifically, LinearRAG constructs a relation-free hierarchical graph, termed Tri-Graph, using only lightweight entity extraction and semantic linking, avoiding unstable relation modeling. This new paradigm of graph construction scales linearly with corpus size and incurs no extra
token consumption, providing an economical and reliable indexing of the original passages. For
retrieval, LinearRAG adopts a two-stage strategy: (i) relevant entity activation via local semantic bridging, followed by (ii) passage
retrieval through global importance aggregation. Extensive experiments on four
datasets demonstrate that LinearRAG significantly outperforms baseline models. Our code and
datasets are available at https://github.com/DEEP-PolyU/LinearRAG.
Read the full abstract ▾
All authors Luyao Zhuang, Shengyuan Chen, Yilin Xiao, Huachi Zhou, Yujing Zhang, Hao Chen, Qinggang Zhang, Xiao Huang
Computation and Language
Stop When Enough: Adaptive Early-Stopping for Chain-of-Thought Reasoning
Renliang Sun, Wei Cheng, Dawei Li et al. (+2 more)
Chain-of-Thought (CoT)
reasoning has driven recent gains of large
language models (LLMs) on
reasoning-intensive tasks by externalizing intermediate steps. However, excessive or redundant
reasoning – so-called overthinking – can increase inference costs and lead LLMs toward incorrect conclusions. In this paper, we present REFRAIN ($\underline{REF}$lective-$\underline{R}$edundancy for $\underline{A}$daptive $\underline{IN}$ference), a training-free framework that adaptively determines when to stop
reasoning to mitigate overthinking. REFRAIN integrates a two-stage stop discriminator to identify reflective yet redundant
reasoning and a sliding-window Upper Confidence Bound (SW-UCB) multi-armed bandit controller to dynamically adjust stopping thresholds according to problem difficulty without supervision or fine-tuning. Across four representative
benchmarks and two model families, REFRAIN reduces
token usage by 20-55% while maintaining or improving accuracy compared to standard CoT prompting. Extensive ablation and robustness analyses demonstrate its stability across models, scorers, and
prompt variations. In summary, our findings highlight when-to-stop as a new and practical axis of test-time scaling – enabling models to reason not just more, but just enough.
Read the full abstract ▾
All authors Renliang Sun, Wei Cheng, Dawei Li, Haifeng Chen, Wei Wang
Computation and Language · Cross-listed from cs.IR
CardRewriter: Leveraging Knowledge Cards for Long-Tail Query Rewriting on Short-Video Platforms
Peiyuan Gong, Feiran Zhu, Yaqi Yin et al. (+6 more)
Short-
video platforms have rapidly become a new generation of information
retrieval systems, where users formulate queries to access desired
videos. However, user queries, especially long-tail ones, often suffer from spelling errors, incomplete phrasing, and ambiguous intent, resulting in mismatches between user expectations and retrieved results. While large
language models (LLMs) have shown success in long-tail query rewriting within e-commerce, they struggle on short-
video platforms, where proprietary content such as short
videos, live streams, micro dramas, and user social networks falls outside their training distribution. To address this challenge, we introduce CardRewriter, an LLM-based framework that incorporates domain-specific knowledge to enhance long-tail query rewriting. For each query, our method aggregates multi-source knowledge relevant to the query and summarizes it into an informative and query-relevant knowledge card. This card then guides the LLM to better capture user intent and produce more effective query rewrites. We optimize CardRewriter using a two-stage training pipeline: supervised fine-tuning followed by group relative policy optimization, with a tailored reward system balancing query relevance and
retrieval effectiveness. Offline experiments show that CardRewriter substantially improves rewriting quality for queries targeting proprietary content. Online A/B testing further confirms significant gains in long-view rate (LVR) and click-through rate (CTR), along with a notable reduction in initiative query reformulation rate (IQRR). Since September 2025, CardRewriter has been deployed on Kuaishou, one of China's largest short-
video platforms, serving hundreds of millions of users daily.
Read the full abstract ▾
All authors Peiyuan Gong, Feiran Zhu, Yaqi Yin, Chenglei Dai, Chao Zhang, Kai Zheng, Wentian Bao, Jiaxin Mao, Yi Zhang
Computation and Language
Diversity Augmentation of Dynamic User Preference Data for Boosting Personalized Text Summarizers
Parthiv Chatterjee, Shivam Sonawane, Amey Hengle et al. (+3 more)
Document summarization enables efficient extraction of user-relevant content but is inherently shaped by individual subjectivity, making it challenging to identify subjective salient information in multifaceted documents. This complexity underscores the necessity for personalized summarization. However, training models for personalized summarization has so far been challenging, particularly because diverse training data containing both user preference history (i.e., click-skip trajectory) and expected (gold-reference) summaries are scarce. The MS/CAS PENS
dataset is a valuable resource but includes only preference history without target summaries, preventing end-to-end supervised learning, and its limited topic-transition diversity further restricts generalization. To address this, we propose $\mathrm{PerAugy}$, a novel cross-trajectory shuffling and summary-content perturbation based data augmentation technique that significantly boosts the accuracy of four state-of-the-art baseline (SOTA) user-encoders commonly used in personalized summarization frameworks (best result: $\text{0.132}$$\uparrow$ w.r.t AUC). We select two such SOTA summarizer frameworks as baselines and observe that when augmented with their corresponding improved user-encoders, they consistently show an increase in personalization (avg. boost: $\text{61.2\%}\uparrow$ w.r.t. PSE-SU4 metric). As a post-hoc analysis of the role of induced diversity in the augmented
dataset by \peraugy, we introduce three
dataset diversity metrics – $\mathrm{TP}$, $\mathrm{RTC}$, and \degreed\ to quantify the induced diversity. We find that $\mathrm{TP}$ and $\mathrm{DegreeD}$ strongly correlate with user-encoder performance on the PerAugy-generated
dataset across all accuracy metrics, indicating that increased
dataset diversity is a key factor driving performance gains.
Read the full abstract ▾
All authors Parthiv Chatterjee, Shivam Sonawane, Amey Hengle, Aditya Tanna, Sourish Dasgupta, Tanmoy Chakraborty
Computation and Language
A-IPO: Adaptive Intent-driven Preference Optimization
Wenqing Wang, Muhammad Asif Ali, Ali Shoker et al. (+4 more)
Human preferences are diverse and dynamic, shaped by regional, cultural, and social factors. Existing
alignment methods like Direct Preference Optimization (DPO) and its variants often default to majority views, overlooking minority opinions and failing to capture latent user intentions in
prompts. To address these limitations, we introduce \underline{A}daptive \textbf{\underline{I}}ntent-driven \textbf{\underline{P}}reference \textbf{\underline{O}}ptimization (A-IPO). Specifically,A-IPO introduces an intention module that infers the latent intent behind each user
prompt and explicitly incorporates this inferred intent into the reward function, encouraging stronger
alignment between the preferred model's responses and the user's underlying intentions. We demonstrate, both theoretically and empirically, that incorporating an intention–response similarity term increases the preference margin (by a positive shift of $\lambda\,\Delta\mathrm{sim}$ in the log-odds), resulting in clearer separation between preferred and dispreferred responses compared to DPO. For evaluation, we introduce two new
benchmarks, Real-pref, Attack-pref along with an extended version of an existing
dataset, GlobalOpinionQA-Ext, to assess real-world and
adversarial preference
alignment. Through explicit modeling of diverse user intents,A-IPO facilitates pluralistic preference optimization while simultaneously enhancing
adversarial robustness in preference
alignment. Comprehensive empirical evaluation demonstrates that A-IPO consistently surpasses existing baselines, yielding substantial improvements across key metrics: up to +24.8 win-rate and +45.6 Response-Intention Consistency on Real-pref; up to +38.6 Response Similarity and +52.2 Defense Success Rate on Attack-pref; and up to +54.6 Intention Consistency Score on GlobalOpinionQA-Ext.
Read the full abstract ▾
All authors Wenqing Wang, Muhammad Asif Ali, Ali Shoker, Ruohan Yang, Junyang Chen, Ying Sha, Huan Wang
Computation and Language
Unilaw-R1: A Large Language Model for Legal Reasoning with Reinforcement Learning and Iterative Inference
Hua Cai, Shuang Zhao, Liang Zhang et al. (+5 more)
Reasoning-focused large
language models (LLMs) are rapidly evolving across various domains, yet their capabilities in handling complex legal problems remains underexplored. In this paper, we introduce Unilaw-R1, a large
language model tailored for legal
reasoning. With a lightweight 7-billion parameter scale, Unilaw-R1 significantly reduces deployment cost while effectively tackling three core challenges in the legal domain: insufficient legal knowledge, unreliable
reasoning logic, and weak business generalization. To address these issues, we first construct Unilaw-R1-Data, a high-quality
dataset containing 17K distilled and screened chain-of-thought (CoT) samples. Based on this, we adopt a two-stage training strategy combining Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL), which significantly boosts the performance on complex legal
reasoning tasks and supports interpretable decision-making in legal AI applications. To assess legal
reasoning ability, we also introduce Unilaw-R1-Eval, a dedicated
benchmark designed to evaluate models across single- and multi-choice legal tasks. Unilaw-R1 demonstrates strong results on authoritative
benchmarks, outperforming all models of similar scale and achieving performance on par with the much larger DeepSeek-R1-Distill-Qwen-32B (54.9%). Following domain-specific training, it also showed significant gains on LawBench and LexEval, exceeding Qwen-2.5-7B-Instruct (46.6%) by an average margin of 6.6%.
Read the full abstract ▾
All authors Hua Cai, Shuang Zhao, Liang Zhang, Xuli Shen, Qing Xu, Weilin Shen, Zihao Wen, Tianke Ban
Computation and Language
CLMN: Concept based Language Models via Neural Symbolic Reasoning
Yibo Yang
Deep learning has advanced NLP, but interpretability remains limited, especially in healthcare and finance. Concept bottleneck models tie predictions to human concepts in
vision, but NLP versions either use binary activations that harm text representations or latent concepts that weaken semantics, and they rarely model dynamic concept interactions such as negation and context. We introduce the Concept
Language Model Network (CLMN), a neural-symbolic framework that keeps both performance and interpretability. CLMN represents concepts as continuous, human-readable
embeddings and applies fuzzy-logic
reasoning to learn adaptive interaction rules that state how concepts affect each other and the final decision. The model augments original text features with concept-aware representations and automatically induces interpretable logic rules. Across multiple
datasets and pre-trained
language models, CLMN achieves higher accuracy than existing concept-based methods while improving explanation quality. These results show that integrating neural representations with symbolic
reasoning in a unified concept space can yield practical, transparent NLP systems.
Read the full abstract ▾
Computation and Language
HUME: Measuring the Human-Model Performance Gap in Text Embedding Tasks
Adnan El Assadi, Isaac Chung, Roman Solomatin et al. (+2 more)
Comparing human and model performance offers a valuable perspective for understanding the strengths and limitations of
embedding models, highlighting where they succeed and where they fail to capture meaning and nuance. However, such comparisons are rarely made, as human performance on
embedding tasks is difficult to measure. To fill this gap, we introduce HUME: Human Evaluation Framework for Text
Embeddings. While frameworks like MTEB provide broad model evaluation, they lack reliable estimates of human performance, limiting the interpretability of model scores. We measure human performance across 16 MTEB
datasets spanning reranking, classification, clustering, and semantic textual similarity across linguistically diverse high- and low-resource
languages. Humans achieve an average performance of 77.6% compared to 80.1% for the best
embedding model, though with substantial variation: models reach high performance on some
datasets while struggling on notably low-resource
languages. Our human annotations also reveal multiple
dataset issues. We additionally
benchmark nine LLMs as annotators on reranking, classification, and STS tasks, finding that they fall short of human performance (76.1% vs. 81.2%) despite offering scalability advantages. We provide human performance baselines, insights into task difficulty patterns, and an extensible evaluation framework that enables a more meaningful interpretation of results and informs the development of both models and
benchmarks. Our code,
dataset, and leaderboard are publicly available at https://github.com/
embeddings-
benchmark/mteb.
Read the full abstract ▾
All authors Adnan El Assadi, Isaac Chung, Roman Solomatin, Niklas Muennighoff, Kenneth Enevoldsen
Computation and Language · Cross-listed from cs.LG
Translution: Unifying Self-attention and Convolution for Adaptive and Relative Modeling
Hehe Fan, Yi Yang, Mohan Kankanhalli et al. (+1 more)
When modeling a given type of data, we consider it to involve two key aspects: 1) identifying relevant elements (e.g., image pixels or textual words) to a central element, as in a convolutional receptive field, or to a query element, as in self-
attention, and 2) encoding these
tokens effectively. Self-
attention can adaptively identify these elements but relies on absolute positional
embedding for structural representation learning. In contrast, convolution encodes elements in a relative manner, yet their fixed kernel size limits their ability to adaptively select the relevant elements. In this paper, we introduce Translution, an operation that unifies the adaptive identification capability of self-
attention and the relative encoding advantage of convolution. However, this integration leads to a substantial increase in the number of parameters, exceeding most currently available computational resources. Therefore, we propose a lightweight variant of Translution, named {\alpha}-Translution. Experiments on computer
vision and natural
language processing tasks show that Translution (including {\alpha}-Translution) achieves superior accuracy compared to self-
attention. The code is available at https://github.com/hehefan/Translution.
Read the full abstract ▾
All authors Hehe Fan, Yi Yang, Mohan Kankanhalli, Fei Wu
Computation and Language
Lightweight Baselines for Medical Abstract Classification: DistilBERT with Cross-Entropy as a Strong Default
Jiaqi Liu, Tong Wang, Su Liu et al. (+4 more)
The research evaluates lightweight medical abstract classification methods to establish their maximum performance capabilities under financial budget restrictions. On the public medical abstracts corpus, we finetune BERT base and Distil BERT with three objectives cross
entropy (CE), class weighted CE, and focal loss under identical tokenization, sequence length, optimizer, and schedule. DistilBERT with plain CE gives the strongest raw argmax trade off, while a post hoc operating point selection (validation calibrated, classwise thresholds) sub stantially improves deployed performance; under this tuned regime, focal benefits most. We report Accuracy, Macro F1, and WeightedF1, release evaluation artifacts, and include confusion analyses to clarify error structure. The practical takeaway is to start with a compact encoder and CE, then add lightweight calibration or thresholding when deployment requires higher macro balance.
Read the full abstract ▾
All authors Jiaqi Liu, Tong Wang, Su Liu, Xin Hu, Ran Tong, Lanruo Wang, Jiexi Xu
Computation and Language
Path Drift in Large Reasoning Models:How First-Person Commitments Override Safety
Yuyi Huang, Runzhe Zhan, Lidia S.Chao et al. (+2 more)
As large
language models (LLMs) are increasingly deployed for complex
reasoning tasks, Long Chain-of-Thought (Long-CoT) prompting has emerged as a key paradigm for structured inference. Despite early-stage safeguards enabled by
alignment techniques such as RLHF, we identify a previously underexplored vulnerability:
reasoning trajectories in Long-CoT models can drift from aligned paths, resulting in content that violates safety constraints. We term this phenomenon Path Drift. Through empirical analysis, we uncover three behavioral triggers of Path Drift: (1) first-person commitments that induce goal-driven
reasoning that delays refusal signals; (2) ethical evaporation, where surface-level disclaimers bypass
alignment checkpoints; (3) condition chain escalation, where layered cues progressively steer models toward unsafe completions. Building on these insights, we introduce a three-stage Path Drift Induction Framework comprising cognitive load amplification, self-role priming, and condition chain hijacking. Each stage independently reduces refusal rates, while their combination further compounds the effect. To mitigate these risks, we propose a path-level defense strategy incorporating role attribution correction and metacognitive reflection (reflective safety cues). Our findings highlight the need for trajectory-level
alignment oversight in long-form
reasoning beyond
token-level
alignment.
Read the full abstract ▾
All authors Yuyi Huang, Runzhe Zhan, Lidia S.Chao, Ailin Tao, Derek F.Wong
Computation and Language
Beyond the limitation of a single query: Train your LLM for query expansion with Reinforcement Learning
Shu Zhao, Tan Yu and Anbang Xu
Reasoning-augmented search
agents, such as Search-R1, are trained to reason, search, and generate the final answer iteratively. Nevertheless, due to their limited capabilities in
reasoning and search, their performance on multi-hop QA
benchmarks remains far from satisfactory. To handle complex or compound queries, we train an LLM-based search
agent with the native capability of query expansion through reinforcement learning. In each turn, our search
agent proposes several query variants, which are searched simultaneously to cover more relevant information. Meanwhile, given limited post-training data and computing resources, it is very challenging for a search
agent to master multiple tasks, including query generation, retrieved information understanding, and answer generation. Therefore, we propose incorporating a pre-trained squeezer model that helps the search
agent understand the retrieved documents, allowing the search
agent to focus on query generation for high
retrieval recall. With the assistance of the squeezer model, we discover that even a small-scale 3B LLM can demonstrate a strong capability of query expansion and achieve state-of-the-art accuracy on the multi-hop QA
benchmarks. To be specific, our experiments across seven question-answering
benchmarks demonstrate that our method, named ExpandSearch, achieves an average improvement of 4.4% compared to state-of-the-art baselines, with strong gains on multi-hop
reasoning tasks requiring diverse evidence aggregation.
Read the full abstract ▾
Computation and Language
MTP-S2UT: Enhancing Speech-to-Speech Translation Quality with Multi-token Prediction
Jianjin Wang, Runsong Zhao, Xiaoqian Liu et al. (+6 more)
Current direct
speech-to-
speech translation methods predominantly employ
speech tokens as intermediate representations. However, a single
speech token is not dense in semantics, so we generally need multiple
tokens to express a complete semantic unit. To address this limitation, we introduce multi-
token prediction (MTP) loss into
speech-to-unit translation (S2UT) models, enabling models to predict multiple subsequent
tokens at each position, thereby capturing more complete semantics and enhancing information density per position. Initial MTP implementations apply the loss at the final layer, which improves output representation but initiates information enrichment too late. We hypothesize that advancing the information enrichment process to intermediate layers can achieve earlier and more effective enhancement of hidden representation. Consequently, we propose MTP-S2UT loss, applying MTP loss to hidden representation where CTC loss is computed. Experiments demonstrate that all MTP loss variants consistently improve the quality of S2UT translation, with MTP-S2UT achieving the best performance.
Read the full abstract ▾
All authors Jianjin Wang, Runsong Zhao, Xiaoqian Liu, Yuan Ge, Ziqiang Xu, Tong Xiao, Shengxiang Gao, Zhengtao Yu, Jingbo Zhu
Computation and Language
Toward Machine Translation Literacy: How Lay Users Perceive and Rely on Imperfect Translations
Yimin Xiao, Yongle Zhang, Dayeon Ki et al. (+5 more)
As Machine Translation (MT) becomes increasingly commonplace, understanding how the general public perceives and relies on imperfect MT is crucial for contextualizing MT research in real-world applications. We present a human study conducted in a public museum (n=452), investigating how fluency and adequacy errors impact bilingual and non-bilingual users' reliance on MT during casual use. Our findings reveal that non-bilingual users often over-rely on MT due to a lack of evaluation strategies and alternatives, while experiencing the impact of errors can
prompt users to reassess future reliance. This highlights the need for MT evaluation and NLP explanation techniques to promote not only MT quality, but also MT literacy among its users.
Read the full abstract ▾
All authors Yimin Xiao, Yongle Zhang, Dayeon Ki, Calvin Bao, Marianna J. Martindale, Charlotte Vaughn, Ge Gao, Marine Carpuat
Computation and Language
Unifying Tree Search Algorithm and Reward Design for LLM Reasoning: A Survey
Jiaqi Wei, Xiang Zhang, Yuejin Yang et al. (+10 more)
Deliberative tree search is a cornerstone of modern Large
Language Model (LLM) research, driving the pivot from brute-force scaling toward algorithmic efficiency. This single paradigm unifies two critical frontiers: Test-Time Scaling (TTS), which deploys on-demand computation to solve hard problems, and Self-Improvement, which uses search-generated data to durably enhance model parameters. However, this burgeoning field is fragmented and lacks a common formalism, particularly concerning the ambiguous role of the reward signal – is it a transient heuristic or a durable learning target? This paper resolves this ambiguity by introducing a unified framework that deconstructs search
algorithms into three core components: the Search Mechanism, Reward Formulation, and Transition Function. We establish a formal distinction between transient Search Guidance for TTS and durable Parametric Reward Modeling for Self-Improvement. Building on this formalism, we introduce a component-centric taxonomy, synthesize the state-of-the-art, and chart a research roadmap toward more systematic progress in creating autonomous, self-improving
agents.
Read the full abstract ▾
All authors Jiaqi Wei, Xiang Zhang, Yuejin Yang, Wenxuan Huang, Juntai Cao, Sheng Xu, Xiang Zhuang, Zhangyang Gao, Muhammad Abdul-Mageed, Laks V.S. Lakshmanan, Chenyu You, Wanli Ouyang, Siqi Sun
Computation and Language · Cross-listed from cs.SE
Operationalizing AI: Empirical Evidence on MLOps Practices, User Satisfaction, and Organizational Context
Stefan Pasch
Organizational efforts to utilize and operationalize artificial intelligence (AI) are often accompanied by substantial challenges, including scalability, maintenance, and coordination across teams. In response, the concept of Machine Learning Operations (MLOps) has emerged as a set of best practices that integrate software engineering principles with the unique demands of managing the ML lifecycle. Yet, empirical evidence on whether and how these practices support users in developing and operationalizing AI applications remains limited. To address this gap, this study analyzes over 8,000 user reviews of AI development platforms from G2.com. Using zero-shot classification, we measure review sentiment toward nine established MLOps practices, including continuous integration and delivery (CI/CD), workflow orchestration, reproducibility, versioning, collaboration, and monitoring. Seven of the nine practices show a significant positive relationship with user satisfaction, suggesting that effective MLOps implementation contributes tangible value to AI development. However, organizational context also matters: reviewers from small firms discuss certain MLOps practices less frequently, suggesting that organizational context influences the prevalence and salience of MLOps, though firm size does not moderate the MLOps-satisfaction link. This indicates that once applied, MLOps practices are perceived as universally beneficial across organizational settings.
Read the full abstract ▾
Computation and Language
Beyond Fertility: Analyzing STRR as a Metric for Multilingual Tokenization Evaluation
Mir Tafseer Nayeem, Sawsan Alqahtani, Md Tahmid Rahman Laskar et al. (+2 more)
Tokenization is a crucial but under-evaluated step in large
language models (LLMs). The standard metric, fertility (the average number of
tokens per word), captures compression efficiency but obscures how vocabularies are allocated across
languages and domains. We analyze six widely used tokenizers across seven
languages and two domains, finding stable fertility for English, high fertility for Chinese, and little domain sensitivity. To address fertility's blind spots, we propose the Single
Token Retention Rate (STRR), which measures the proportion of words preserved as single
tokens. STRR reveals systematic prioritization of English, strong support for Chinese, and fragmentation in Hindi, offering an interpretable view of cross-lingual
fairness. Our results show that STRR complements fertility and provides practical guidance for designing more equitable
multilingual tokenizers.
Read the full abstract ▾
All authors Mir Tafseer Nayeem, Sawsan Alqahtani, Md Tahmid Rahman Laskar, Tasnim Mohiuddin, M Saiful Bari
Computation and Language
Unpacking Hateful Memes: Presupposed Context and False Claims
Weibin Cai, Jiayu Li and Reza Zafarani
While memes are often humorous, they are frequently used to disseminate hate, causing serious harm to individuals and society. Current approaches to hateful meme detection mainly rely on pre-trained
language models. However, less focus has been dedicated to what make a meme hateful. Drawing on insights from philosophy and psychology, we argue that hateful memes are characterized by two essential features: a presupposed context and the expression of false claims. To capture presupposed context, we develop PCM for modeling contextual information across modalities. To detect false claims, we introduce the FACT module, which integrates external knowledge and harnesses cross-modal reference graphs. By combining PCM and FACT, we introduce \textbf{SHIELD}, a hateful meme detection framework designed to capture the fundamental nature of hate. Extensive experiments show that SHIELD outperforms state-of-the-art methods across
datasets and metrics, while demonstrating versatility on other tasks, such as fake news detection.
Read the full abstract ▾