Summarization
-
Improving Multi-turn Dialogue Consistency with Self-Recall Thinking
2026-05-14 17:20:14Renning Pang, Tian Lan, Leyuan Liu, Xiaoming Huang, Piao Tong, Xiaosong ZhangAbstract
Large language model (LLM) based multi-turn dialogue systems often struggle to track dependencies across non-adjacent turns, undermining both consistency and scalability. As conversations lengthen, essential information becomes sparse and is buried in irrelevant context, while processing the entire dialogue history incurs severe efficiency bottlenecks. Existing solutions either rely on high latency external memory or lose fine-grained details through iterative summarization. In this paper, we propose Self-Recall Thinking (SRT), a framework designed to address long-range contextual dependency and sparse informative signals in multi-turn dialogue. SRT identifies helpful historical turns and uses them to generate contextually appropriate responses, enabling the model to selectively recall and reason over context during inference. This process yields an endogenous reasoning process that integrates interpretable recall steps without external modules. SRT incorporates: (1) Dependency Construction: Generating and converting it into self-recall chains; (2)Capability Initialization: Training to enable reasoning chains with recall tokens capability; (3)Reasoning Improvement: Refining accuracy via verifiable rewards to optimize recall and reasoning for correct answers. Experiments on multiple datasets demonstrate that SRT improves F1 score by 4.7% and reduces end-to-end latency by 14.7% over prior methods, achieving a balance between reasoning latency and accuracy, and outperforming state-of-the-art baselines.
Abstract (translated)
URL
https://arxiv.org/abs/2605.15102
PDF
-
COREKG: Coreset-Guided Personalized Summarization of Knowledge Graphs
2026-05-14 14:40:05Sohel Aman Khan, Raghava Mutharaju, Supratim ShitAbstract
Knowledge Graphs (KGs) are extensively used across different domains and in several applications. Often, these KGs are very large in size. Such KGs become unwieldy for tasks such as question answering and visualization. Summarization of KGs offers a viable alternative in such cases. Furthermore, personalized KG summarization is crucial in the current data-driven world as it captures the specific requirements of users based on their query patterns. Since it only maintains relevant information, the personalized summaries of KG are small, resulting in significantly smaller storage requirements and query runtime. In this work, we adapt the coreset theory to create personalized KG summaries. For a given dataset and a user-specific query workload, we present an approach that samples a relevant subset of triples using sensitivity-based importance sampling. We ensure that the subset approximates the characteristics of the full dataset with bounded approximation error. We define sensitivity scores that measure the importance of a triple with respect to a user's query workload, which are then used by our coreset construction algorithm. We explicitly focus on personalized knowledge graph summarization by constructing summaries independently for each user based on their query behaviour. Our evaluation on Freebase, WikiData, and DBpedia shows that COREKG delivers higher query-answering accuracy and structural coverage than the state-of-the-art methods, such as GLIMPSE, PPR, iSummary, PEGASUS and APEX$^2$ while requiring only a tiny fraction of the original graph.
Abstract (translated)
URL
https://arxiv.org/abs/2605.14900
PDF
-
Herculean: An Agentic Benchmark for Financial Intelligence
2026-05-14 04:30:49Xueqing Peng, Zhuohan Xie, Yupeng Cao, Haohang Li, Lingfei Qian, Yan Wang, Vincent Jim Zhang, Huan He, Xuguang Ai, Linhai Ma, Ruoyu Xiang, Yueru He, Yi Han, Shuyao Wang, Yuqing Guo, Mingyang Jiang, Yilun Zhao, Youzhong Dong, Xiaoyu Wang, Yankai Chen, Ye Yuan, Qiyuan Zhang, Fuyuan Lyu, Haolun Wu, Yonghan Yang, Zichen Zhao, Yuyang Dai, Fan Zhang, Rania Elbadry, Ayesha Gull, Muhammad Usman Safder, Nuo Chen, Fengbin Zhu, Tianshi Cai, Zimu Wang, Polydoros Giannouris, Yuechen Jiang, Zhiwei Liu, Mohsinul Kabir, Yuyan Wang, Yixiang Zheng, Yangyang Yu, Weijin Liu, Wenbo Cao, Anke Xu, Peng Lu, Jerry Huang, Fengran Mo, Mingquan Lin, Prayag Tiwari, Yijia Zhao, Victor Gutierrez Basulto, Xiao-Yang Liu, Kaleb E Smith, Jiahuan Pei, Arman Cohan, Jimin Huang, Yuehua Tang, Alejandro Lopez-Lira, Xi Chen, Xue Liu, Junichi Tsujii, Jian-Yun Nie, Sophia AnaniadouAbstract
As AI agents improve, the central question is no longer whether they can solve isolated well-defined financial tasks, but whether they can reliably carry out financial professional work. Existing financial benchmarks offer only a partial view of this ability, as they primarily evaluate static competencies such as question answering, retrieval, summarization, and classification. We introduce Herculean, the first skilled benchmark for agentic financial intelligence spanning four representative workflows, including Trading, Hedging, Market Insights, and Auditing. Each workflow is instantiated as a standardized MCP-based skill environment with its own tools, interaction dynamics, constraints, and success criteria, enabling consistent end-to-end assessment of heterogeneous agent systems. Across frontier agents, we find agents perform relatively well on Trading and Market Insights, but struggle substantially on Hedging and Auditing, where long-horizon coordination, state consistency, and structured verification are critical. Overall, our results point to a key gap in current agents in turning financial reasoning into dependable workflow execution in high-stakes financial workflows.
Abstract (translated)
URL
https://arxiv.org/abs/2605.14355
PDF
-
BOOKMARKS: Efficient Active Storyline Memory for Role-playing
2026-05-13 22:48:24Letian Peng, Ziche Liu, Yiming Huang, Longfei Yun, Kun Zhou, Yupeng Hou, Jingbo ShangAbstract
Memory systems are critical for role-playing agents (RPAs) to maintain long-horizon consistency. However, existing RPA memory methods (e.g., profiling) mainly rely on recurrent summarization, whose compression inevitably discards important details. To address this issue, we propose a search-based memory framework called BOOKMARKS, which actively initializes, maintains, and updates task-relevant pieces of bookmarks for the current task (e.g., character acting). A bookmark is structured as the answer to a question at a specific point in the storyline. For each current task, BOOKMARKS selects reusable existing bookmarks or initializes new ones (at storyline beginning) with useful questions. These bookmarks are then synchronized to the current story point, with their answers updated accordingly, so they can be efficiently reused in future grounding rounds. Compared with recurrent summarization, BOOKMARKS offers (1) active grounding for capturing task-specific details and (2) passive updating to avoid unnecessary computation. In implementation, BOOKMARKS supports concept, behavior, and state searches, each powered by an efficient synchronization method. BOOKMARKS significantly outperforms RPA memory baselines on 85 characters from 16 artifacts, demonstrating the effectiveness of search-based memory for RPAs.
Abstract (translated)
URL
https://arxiv.org/abs/2605.14169
PDF
-
A Non-Destructive Methodological Framework for Modernizing Legacy Clinical Reporting Systems for AI-Driven Pharmacoinformatics: A SAS Case Study
Abstract
Drug development and pharmacovigilance are frequently bottlenecked by legacy clinical reporting pipelines. These monolithic systems encode regulatory-grade logic but resist AI integration by producing opaque output with no machine-readable intermediate layer. Existing modernization approaches force a choice between full rewrites and incremental refactoring that preserves structural barriers. We present a non-destructive methodological framework achieving AI-driven pharmacoinformatics readiness without altering legacy source code. A metadata layer--comprising a bridge map, a typed Intermediate Representation (IR), and an orchestrator--wraps existing components and re-exposes their outputs as structured data consumable by LLMs. It enables optional incremental consolidation, replacing selected legacy components with metadata-configured core routines while the remainder operates unchanged. Validated on a 558-component SAS reporting library (373,000 lines of code), the framework demonstrated immediate AI-readiness under coexistence mode, yielding machine-readable output. Where consolidation was elected, the modernized core achieved a 92% reduction in proprietary code. Parity validation on 14 report types from a Phase III study achieved cell-level parity of 80% or above on 11 reports (mean 82.7%, best 99.2%). A benchmark using CDISC CDISCPilot01 data achieved 100% parity across 5 reports. LLM experiments confirmed the IR enables automated pharmacovigilance, table summarization, and trial configuration generation. The framework offers a regulation-aware path to AI-integrated clinical reporting, accelerating drug development without interrupting regulatory submissions.
Abstract (translated)
URL
https://arxiv.org/abs/2605.13905
PDF
-
TokenRatio: Principled Token-Level Preference Optimization via Ratio Matching
2026-05-12 15:44:33Truong Nguyen, Tien-Phat Nguyen, Linh Ngo Van, Duy Minh Ho Nguyen, Khoa Doan, Trung LeAbstract
Direct Preference Optimization (DPO) is a widely used RL-free method for aligning language models from pairwise preferences, but it models preferences over full sequences even though generation is driven by per-token decisions. Existing token-level extensions typically decompose a sequence-level Bradley-Terry objective across timesteps, leaving per-prefix (state-wise) optimality implicit. We study how to recover token-level preference optimality using only standard sequence-level pairwise comparisons. We introduce Token-level Bregman Preference Optimization (TBPO), which posits a token-level Bradley-Terry preference model over next-token actions conditioned on the prefix, and derive a Bregman-divergence density-ratio matching objective that generalizes the logistic/DPO loss while preserving the optimal policy induced by the token-level model and maintaining DPO-like simplicity. We introduce two instantiations: TBPO-Q, which explicitly learns a lightweight state baseline, and TBPO-A, which removes the baseline through advantage normalization. Across instruction following, helpfulness/harmlessness, and summarization benchmarks, TBPO improves alignment quality and training stability and increases output diversity relative to strong sequence-level and token-level baselines.
Abstract (translated)
URL
https://arxiv.org/abs/2605.12288
PDF
-
Towards Visually-Guided Movie Subtitle Translation for Indic Languages
2026-05-12 11:43:57Tarun Chintada, Kshetrimayum Boynao Singh, Asif EkbalAbstract
Movie subtitle translation is inherently multimodal, yet text-only systems often miss visual cues needed to convey emotion, action, and social nuance, especially for low-resource Indic languages (English to Hindi, Bengali, Telugu, Tamil and Kannada). We present a case study on five full-length films and compare two lightweight visual grounding strategies: structured attribute summaries from a 5-minute sliding window and free-text summaries of inter-subtitle visual gaps. Our analysis shows that temporal misalignment between subtitles and frames is a major obstacle in long-form video, often rendering indiscriminate visual grounding ineffective. However, oracle selective grounding, which replaces only the lowest-quality 20-30\% of baseline segments with visual-enhanced outputs, consistently improves COMET over the text-only baseline while requiring far less visual processing. Among the two approaches, coarse attribute-based visual context summarization is more robust, capturing scene-level emotion and contextual subtle cues that text alone often misses
Abstract (translated)
URL
https://arxiv.org/abs/2605.11993
PDF
-
Multimodal Abstractive Summarization of Instructional Videos with Vision-Language Models
2026-05-12 11:11:36Maham Nazir, Muhammad Aqeel, Richong Zhang, Francesco SettiAbstract
Multimodal video summarization requires visual features that align semantically with language generation. Traditional approaches rely on CNN features trained for object classification, which represent visual concepts as discrete categories not aligned with natural language. We propose ClipSum, a framework that leverages frozen CLIP vision-language features with explicit temporal modeling and dimension-adaptive fusion for instructional video summarization. CLIP's contrastive pre-training on 400M image-text pairs yields visual features semantically aligned with the linguistic concepts that text decoders generate, bridging the vision-language gap at the representation level. On YouCook2, ClipSum achieves 33.0% ROUGE-1 versus 30.5% for ResNet-152 with 4x lower dimensionality (512 vs. 2048), demonstrating that semantic alignment matters more than feature capacity. Frozen CLIP (33.0%) surpasses fine-tuned CLIP (32.3%), showing that preserving pre-trained alignment is more valuable than task-specific adaptation. this https URL
Abstract (translated)
URL
https://arxiv.org/abs/2605.11959
PDF
-
GridProbe: Posterior-Probing for Adaptive Test-Time Compute in Long-Video VLMs
2026-05-11 15:57:46Mohamed Eltahir, Lama Ayash, Ali Habibullah, Tanveer Hussain, Naeemullah KhanAbstract
Long-video understanding in VLMs is bottlenecked by a single monolithic forward pass over thousands of frames at quadratic attention cost. A common mitigation is to first select a small subset of informative frames before the forward pass; common for training-free selectors via auxiliary encoder-space similarities. Such signals are capped by contrastive pretraining, which usually fails on reasoning-heavy queries (negation, cross-frame counting, holistic summarization). We propose GridProbe, an efficient training-free posterior-probing inference paradigm that scores evidence in answer space using a frozen VLM's own reasoning and then selects question-relevant frames adaptively, resulting in sub-quadratic attention cost with little to no accuracy loss. We arrange frames on a $K{\times}K$ grid and run lightweight row R and column C probes, where each probe reads its peak posterior as a query-conditioned confidence. The outer product of R and C yields an interpretable importance map whose skewness and kurtosis drive Shape-Adaptive Selection, a closed-form rule that reliably replaces the fixed frame budget $M$ with a per-question $M_{\mathrm{eff}}$. We show empirically that $M_{\mathrm{eff}}$ tracks intrinsic question difficulty without ever seeing the answer, a sign of test-time adaptive compute. On Video-MME-v2, GridProbe matches the monolithic baseline within $1.6$ pp Avg Acc at $3.36\times$ TFLOPs reduction, while on LongVideoBench it Pareto-dominates the baseline ($+0.9$ pp at $0.35\times$ compute). Because the selector and QA models can be decoupled, pairing a small 2B selector with a stronger 4B or 8B QA is strictly Pareto-dominant over the 2B monolithic baseline (up to $+4.0$ pp at $0.52\times$ compute, on average), with no retraining. Finally, the interpretability of the importance maps opens future avenues for behavioral diagnostics, grounding, and frame-selection distillation.
Abstract (translated)
URL
https://arxiv.org/abs/2605.10762
PDF
-
Navigating the Sea of LLM Evaluation: Investigating Bias in Toxicity Benchmarks
2026-05-11 14:27:39Regina Gugg, Selina Niederl\"ander, Andreas St\"ockl, Martin FlechlAbstract
The rapid adoption of LLMs in both research and industry highlights the challenges of deploying them safely and reveals a gap in the systematic evaluation of toxicity benchmarks. As organizations increasingly rely on these benchmarks to certify models for customer-facing applications and automated moderation, unrecognized evaluation biases could lead to the deployment of vulnerable or unsafe systems. This work investigates the robustness of established benchmarking setups and examines how to measure currently neglected intrinsic biases, such as those related to model choice, metrics, and task types. Our experiments uncover significant discrepancies in benchmark behaviors when evaluation setups are altered. Specifically, shifting the task from text completion to summarization increases the tendency of benchmarks to flag content as harmful. Additionally, certain benchmarks fail to maintain consistent behavior when the input data domain is changed. Furthermore, we observe model-specific instabilities, demonstrating a clear need for more robust and comprehensive safety evaluation frameworks.
Abstract (translated)
URL
https://arxiv.org/abs/2605.10639
PDF
-
NyayaAI: An AI-Powered Legal Assistant Using Multi-Agent Architecture and Retrieval-Augmented Generation
2026-05-11 08:04:07Deepanshu, Divi Saxena, Deepali Rana, Ayesha Varshney, Sahinur Rahman LaskarAbstract
Legal information in India remains largely inaccessible due to the complexity of legal language and the sheer volume of legal documentation involved in research and case analysis. This paper presents NyayaAI, an AI-powered legal assistant that automates and simplifies legal workflows for lawyers, law students, and general users. The system combines Large Language Models with a Retrieval-Augmented Generation pipeline grounded in a curated Indian legal knowledge base comprising constitutional provisions, statutes, case laws, and judicial precedents. A multi-agent architecture orchestrated through the Mastra TypeScript framework coordinates a main agent with specialized sub-agents handling legal research, document summarization, case law retrieval, and drafting assistance. A compliance module validates all responses before delivery. Domain classification achieved 70\% precision across test samples, with RAG retrieval precision at 74\% and overall response accuracy at 72\%, demonstrating that structured multi-agent LLM systems can meaningfully improve legal accessibility and workflow efficiency. The code\footnote{this https URL} is made publicly available for the benefit of the research community.
Abstract (translated)
URL
https://arxiv.org/abs/2605.10155
PDF
-
EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding
2026-05-11 01:59:59Ziyang Wang, Yue Zhang, Shoubin Yu, Ce Zhang, Zengqi Zhao, Jaehong Yoon, Hyunji Lee, Gedas Bertasius, Mohit BansalAbstract
Next-generation visual assistants, such as smart glasses, embodied agents, and always-on life-logging systems, must reason over an entire day or more of continuous visual experience. In ultra-long video settings, relevant information is sparsely distributed across hours or days, making memory a fundamental challenge: models must accumulate information over time, recall prior states, track temporal order, and abstract recurring patterns. However, existing week-long video benchmarks are primarily designed for perception and recognition, such as moment localization or global summarization, rather than reasoning that requires integrating evidence across multiple days. To address this gap, we introduce EgoMemReason, a comprehensive benchmark that systematically evaluates week-long egocentric video understanding through memory-driven reasoning. EgoMemReason evaluates three complementary memory types: entity memory, tracking how object states evolve and change across days; event memory, recalling and ordering activities separated by hours or days; and behavior memory, abstracting recurring patterns from sparse, repeated observations over the whole week period. EgoMemReason comprises 500 questions across three memory types and six core challenges, with an average of 5.1 video segments of evidence per question and 25.9 hours of memory backtracking. We evaluate EgoMemReason on 17 methods across MLLMs and agentic frameworks, revealing that even the best model achieves only 39.6% overall accuracy. Further analysis shows that the three memory types fail for distinct reasons and that performance degrades as evidence spans longer temporal horizons, revealing that long-horizon memory remains far from solved. We believe EgoMemReason establishes a strong foundation for evaluating and advancing long-context, memory-aware multimodal systems.
Abstract (translated)
URL
https://arxiv.org/abs/2605.09874
PDF
-
Insight: Enhancing Mobile Accessibility for Blind and Visually Impaired Users with LLMs
2026-05-10 22:51:44Joshua Owusu Ansah, Anuj Kapoor, Ayush Khanna, Manvika Vinod, Precious Njeck, Shuai GaoAbstract
This research paper addresses the limitations of current mobile accessibility services like TalkBack, which provide manual gesture-based sequential feedback to BVI users. Motivated by the promise of large language models (LLMs), this paper introduces Insight, an Android accessibility service that provides natural language interaction and real-time summarization of the screen. The paper performs a within-subject experimental study with users to compare Insight and TalkBack on usability factors. Results show Insight reduced mental effort and task time, and was preferred because of its dialogue interface, but users felt the need for interruption management. Results show LLM-based interfaces can significantly improve mobile accessibility, and describe the potential of hybrid solutions combining gesture and dialogue modalities towards more inclusive design.
Abstract (translated)
URL
https://arxiv.org/abs/2605.09803
PDF
-
Uncertainty-Aware and Decoder-Aligned Learning for Video Summarization
2026-05-10 12:31:02Omer Tariq, Syed Muhammad Raza, Jeongbae SonAbstract
Video summarization aims to produce a compact representation of a long video by selecting a subset of temporally important segments that best reflect human preferences. This task is inherently difficult due to strong annotation subjectivity and the reliance on discrete decoding procedures, such as temporal segmentation and knapsack-based selection, during evaluation. Most existing approaches either learn deterministic importance scores that overlook these characteristics or adopt complex generative models that increase training and inference cost. In this paper, we propose VASTSum, an uncertainty-aware and decoder-aligned learning framework for video summarization that addresses both challenges within a single-pass model. The proposed method predicts probabilistic frame-level importance scores using a variational formulation, enabling explicit modeling of uncertainty arising from multi-annotator supervision. To account for subjectivity, particularly under binary annotations, we employ a supervision strategy that encourages alignment with plausible human annotation modes rather than enforcing a single consensus target. Furthermore, we introduce a decoder-aligned regularization that promotes stability of knapsack-based summary selection, reducing sensitivity to small perturbations in predicted scores. We evaluate the proposed framework on the SumMe and TVSum benchmarks using standard rank-based metrics. Experimental results show consistent and competitive Kendall and Spearman correlations across multiple data splits, demonstrating improved robustness under annotation disagreement while maintaining efficient single-forward inference. These results indicate that explicitly modeling uncertainty and aligning learning objectives with the decoding stage provide a principled alternative to both deterministic and diffusion-based video summarization methods.
Abstract (translated)
URL
https://arxiv.org/abs/2605.09507
PDF
-
Don't Click That: Teaching Web Agents to Resist Deceptive Interfaces
2026-05-10 12:11:52Yilin Zhang, Yingkai Hua, Chunyu Wei, Xin Wang, Yueguo ChenAbstract
Vision-language model (VLM) based web agents demonstrate impressive autonomous GUI interaction but remain vulnerable to deceptive interface elements. Existing approaches either detect deception without task integration or document attacks without proposing defenses. We formalize deception-aware web agent defense and propose DUDE (Deceptive UI Detector & Evaluator), a two-stage framework combining hybrid-reward learning with asymmetric penalties and experience summarization to distill failure patterns into transferable guidance. We introduce RUC (Real UI Clickboxes), a benchmark of 1,407 scenarios spanning four domains and deception categories. Experiments show DUDE reduces deception susceptibility by 53.8% while maintaining task performance, establishing an effective foundation for robust web agent deployment.
Abstract (translated)
URL
https://arxiv.org/abs/2605.09497
PDF
-
Do Benchmarks Underestimate LLM Performance? Evaluating Hallucination Detection With LLM-First Human-Adjudicated Assessment
2026-05-08 20:27:44I. F. Atasoy, B. Mutlu, E. A. Sezer, A. WahdanAbstract
Hallucination remains a persistent challenge in Large Language Models (LLMs), particularly in context-grounded settings such as RAG and agentic AI systems. This study focuses on contextual hallucination detection in summarization tasks. We analyze the QAGS-C and SummEval datasets by comparing original benchmark annotations with reason and span-based predictions from Gemini 2.5 Flash and GPT-5 Mini. To address systematic divergences between human labels and LLM judgments, we re-evaluated all conflicted samples through a human adjudication process involving 2 cross-cultural adjudicators. Following this re-evaluation, triple agreement (between human, GPT, and Gemini) increased by 6.38% for QAGS-C and 7.62% for SummEval. Similarly, model accuracy improved, with GPT increasing by 4.25% on QAGS-C and 2.34% on SummEval, while Gemini showed gains of 8.51% and 3.80%, respectively. Notably, adjudicators frequently sided with the models' judgments over original human annotations when LLMs provided explicit reasoning. Overall human adjudicator agreement ranged between 83% and 87%. These findings suggest that for ambiguity-prone tasks, single-pass annotations may be insufficient, and model-assisted re-evaluation yields more reliable benchmarks.
Abstract (translated)
URL
https://arxiv.org/abs/2605.08462
PDF
-
Theoretical Limits of Language Model Alignment
2026-05-08 01:32:22Lucas Monteiro Paes, Natalie Mackraz, Barry-John Theobald, Federico DanieliAbstract
Language model (LM) alignment improves model outputs to reflect human preferences while preserving the capabilities of the base model. The most common alignment approaches are (i) reinforcement learning, which maximizes the expected reward under a KL-divergence constraint, and (ii) best-of-$N$ alignment, which selects the highest-reward output among $N$ independent samples. Despite their widespread use, the fundamental limits of reward improvement under a KL budget remain poorly understood. We characterize the information-theoretic limits of KL-regularized alignment by deriving the maximum achievable expected reward gain for a fixed KL-divergence budget. Our first result provides a closed-form expression for the optimal reward improvement, governed by a Jeffreys divergence term rather than the $\sqrt{\texttt{KL}}$ used in prior analyses. We further reformulate this expression as a covariance under the base model, yielding a practical estimator that predicts achievable alignment gains from base model samples alone. We extend our analysis to the proxy reward setting, showing that the gap between ideal and proxy alignment (reward hacking) grows with the magnitude of reward error and when the KL penalty factor decreases. We then prove that reward ensembling mitigates reward hacking, providing a theoretical justification for this technique used in practice. Empirically, we compute the KL-reward Pareto frontier for two tasks for LMs, safety and summarization, and show that best-of-$N$ closely approaches the theoretical limit, while PPO and GRPO remain substantially suboptimal. Our theoretical results shed light on several empirically observed phenomena in the alignment literature and suggest that algorithmic improvements are needed to achieve optimal alignment without high inference costs.
Abstract (translated)
URL
https://arxiv.org/abs/2605.07105
PDF
-
Self-Consolidating Language Models: Continual Knowledge Incorporation from Context
2026-05-08 00:50:09Zekun Wang, Anant Gupta, Zihan Dong, Christopher J. MacLellanAbstract
Large language models (LLMs) increasingly receive information as streams of passages, conversations, and long-context workflows. While longer context windows expose more evidence, they do not ensure that useful information is preserved and reused. We study continual context consolidation: writing current context into model weights while limiting interference with previously consolidated information. We propose \textbf{S}elf-\textbf{Co}nsolidating \textbf{L}anguage Models (SCoL), a post-training framework in which, given current context, an LLM learns to generate textual update instructions specifying which of its own Transformer layers should be updated. Because committed updates change the model that later generates future selections, we train SCoL with meta-reinforcement learning over an evolving model state. We instantiate SCoL with supervised QA rewards on SQuAD knowledge incorporation and intrinsic likelihood-based rewards for LongBench v2 long-context consolidation. Across both settings, SCoL improves acquisition and retention over prompting, summarization, batch test-time training, and sequential finetuning baselines. Analysis of learned selection patterns shows that SCoL encourages the LLM to generate sparse update locations that align with layers of high Fisher information, suggesting that the model learns to route plasticity toward loss-sensitive regions while limiting interference. Moreover, SCoL transfers from shorter meta-training streams to longer LongBench v2 streams at evaluation, suggesting that our framework supports scalable streaming consolidation.
Abstract (translated)
URL
https://arxiv.org/abs/2605.07076
PDF
-
Human-AI Co-Evolution and Epistemic Collapse: A Dynamical Systems Perspective
2026-05-07 14:31:17Xuening Wu, Yanlan Kang, Qianya Xu, Kexuan Xie, Jiaqi Mi, Honggang Wang, Yubin Liu, Zeping ChenAbstract
Large language models (LLMs) are reshaping how knowledge is produced, with increasing reliance on AI systems for generation, summarization, and reasoning. While prior work has studied cognitive offloading in humans and model collapse in recursive training, these effects are typically considered in isolation. We propose a unified perspective: humans and language models form a coupled dynamical system linked by a feedback loop of usage, generation, and retraining. We introduce a minimal model with three variables -- human cognition, data quality, and model capability -- and show that this feedback can give rise to distinct dynamical regimes. Our analysis identifies three regimes: co-evolutionary enhancement, fragile equilibrium, and degenerative convergence. Through a simple simulation, we demonstrate that increasing reliance on AI can induce a transition toward a low-diversity, suboptimal equilibrium. From an information-theoretic perspective, this transition corresponds to an emergent information bottleneck in the human-AI loop, where entropy reduction reflects loss of diversity and support under closed-loop feedback rather than beneficial compression. These results suggest that the trajectory of AI systems is shaped not only by model design, but by the dynamics of human-AI co-evolution.
Abstract (translated)
URL
https://arxiv.org/abs/2605.06347
PDF
-
PREFER: Personalized Review Summarization with Online Preference Learning
2026-05-07 09:20:47Millend Roy, Agostino Capponi, Vineet GoyalAbstract
Product reviews significantly influence purchasing decisions on e-commerce platforms. However, the sheer volume of reviews can overwhelm users, obscuring the information most relevant to their specific needs. Current e-commerce summarization systems typically produce generic, static summaries that fail to account for the fact that (i) different users care about different product characteristics, and (ii) these preferences may evolve with interactions. To address the challenge of unknown latent preferences, we propose an online learning framework that generates personalized summaries for each user. Our system iteratively refines its understanding of user preferences by incorporating feedback directly from the generated summaries over time. We provide a case study using the Amazon Reviews'23 dataset, showing in controlled simulations that online preference learning improves alignment with target user interests while maintaining summary quality.
Abstract (translated)
URL
https://arxiv.org/abs/2605.05911
PDF
