Search: “arxiv-cs-cl”

Layer-Informed Fine-Tuning via Three-Stage Functional Segmentation of LLMs

arXiv:2609.38027v1 Announce Type: new Abstract: In recent years, the performance of large language models (LLMs) on reasoning tasks has been remarkable, even surpassing human capabilities on various b

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Learning from Teacher Continuations at Student States

DGX agent

arXiv:2609.36246v1 Announce Type: new Abstract: We present OLIVE (OnLine InterVEntion). At each iteration, the evolving student policy generates a new prefix, the teacher continues it autoregressively

model-releasesarxiv-cs-cl

30 Sept 2026

Learning from Think-Mode Advantage via On-Policy Distillation

arXiv:2609.37044v1 Announce Type: new Abstract: Explicit intermediate reasoning gives large language models (LLMs) a stronger problem-solving mode. We study learning from this think-mode advantage via

Learning What to Remember: Long-horizon Counterfactual Memory Optimization

arXiv:2609.37930v1 Announce Type: new Abstract: Persistent textual memory allows language models to carry information across long interactions, but learning what to remember is fundamentally a credit-

LLM unbranding: Erasing Commercial Identity while Preserving Generic Utility

arXiv:2609.37127v1 Announce Type: new Abstract: Establishing unbranding as a critical practice to prevent visual logos from acquiring negative connotations is standard in image generation. Large Langu

Tutorials

LLMs learn different forms of metacognition when trained to predict their own accuracy

arXiv:2609.33886v2 Announce Type: replace Abstract: Large language models are trained to always produce an answer, regardless of whether they possess the relevant knowledge, which leads them to fabric

tutorialsarxiv-cs-cl

LoLBench: Evaluating Coding Agents with Long-Horizon Proposals on Large Software Systems

arXiv:2609.37143v1 Announce Type: cross Abstract: Modern coding agents can deliver increasingly large repository-level changes, and recent benchmarks reflect this by emphasizing long-horizon tasks wit

LongHarness Bench: Stress-Testing Language Model Harnesses for Long-Context Reasoning

arXiv:2609.38137v1 Announce Type: new Abstract: Language-model (LM) harnesses enable LMs to operate effectively over long contexts using additional compute. However, existing long-context evaluations

Look What You Made Us Cluster: Hate Narrative Extraction from Reddit Discourse

arXiv:2609.37408v1 Announce Type: new Abstract: Narrative extraction allows us to identify online hate narratives, supporting the construction of rigorous detection systems. Existing computational app

Lost in Translation: Measuring the Effect of Non-Native English on End User Performance of Large Language Models

arXiv:2609.36214v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used by people whose first language is not English, yet these users have been shown to receive systematica

MAPLE: Medical Aspect-Based Summarization with Phrase-Level Evidence

arXiv:2601.03418v3 Announce Type: replace Abstract: Trustworthy clinical summarization requires every claim to be traceable to its evidence, yet existing attribution often resolves only to the sentenc

MARCO: Multi-Round Agentic Reinforcement for Conditional Molecular Optimization

arXiv:2609.36683v1 Announce Type: cross Abstract: Molecular optimization is inherently iterative: a candidate is proposed, evaluated against several objectives, and revised while preserving a relation

MemFold: Learning Compact Soft Memory for Long-Context Personalization via On-Policy Optimization

arXiv:2609.36435v1 Announce Type: new Abstract: An assistant that serves the same user over a long horizon has to answer from what that user has revealed: which preferences still hold, which were revi

Memory Consolidation Flattens the Temporal Shape of User Facts

arXiv:2609.36457v1 Announce Type: new Abstract: Long-term memory systems turn conversations into short stored notes. A note can keep a user fact while losing evidence about whether the fact still hold

MERGE: Multi-LLM Ensemble for Retrieval via Generative Enrichment

arXiv:2609.37574v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used to enrich user queries in information retrieval (IR) so that a standard retriever such as BM25 can

Momentum-Coupled Rubric Adaptation for Detailed Image Captioning

arXiv:2609.36893v1 Announce Type: new Abstract: Detailed image captioning requires accurate and comprehensive descriptions of fine-grained visual content, yet caption quality spans factual accuracy, i

MONOVAB : An Annotated Corpus for Bangla Multi-label Emotion Detection

arXiv:2309.15670v3 Announce Type: replace-cross Abstract: In recent years, Sentiment Analysis (SA) and Emotion Recognition (ER) have been increasingly popular in the Bangla language, which is the seve

Tutorials

Multimodal Detection of Higher-Order Behavioral Constructs: Self-Compassion in Structured Reflective Interaction

arXiv:2609.37148v1 Announce Type: cross Abstract: Many of the qualities that matter most in how people learn and grow, how someone regulates their emotions, reflects on a setback, or stays aware of ot

tutorialsarxiv-cs-cl

One Threshold Does Not Fit All Languages: Language-Conditional Deferral for Reliable and Efficient Low-Resource Text Classification

arXiv:2609.37861v1 Announce Type: new Abstract: In the Global South, the lower-income countries of Africa, Asia, and Latin America where most of the world's languages are spoken, a deployed text class

Orthogonal Yet Coupled: Decoupling Geometric Components for Model Merging

arXiv:2609.37564v1 Announce Type: new Abstract: Merging pretrained models has emerged as an effective approach for consolidating diverse capabilities into a single unified model. However, prevailing m

Overcoming Scaling Limits in On-Policy Self-Distillation for LLM Reasoning

arXiv:2609.37915v1 Announce Type: cross Abstract: On-policy self-distillation (OPSD) trains a student to match a privileged teacher distribution along its own sampled trajectory. Standard OPSD applies

Pair Difficulty Matters: Rethinking Pairwise LLM-as-a-Judge Evaluation and Consistency

arXiv:2609.37577v1 Announce Type: new Abstract: Large Language Model judges are widely used to rank texts and text-generating systems through pairwise comparison, and their reliability is typically as

POET: Preference Optimization for Enhanced Text-to-Image Generation

arXiv:2510.12041v3 Announce Type: replace Abstract: Recent advances in text-to-image (T2I) generation have achieved impressive results, yet existing models often struggle with simple or underspecified

Predicting Team Performance from Communications in Simulated Search-and-Rescue

arXiv:2503.03791v1 Announce Type: cross Abstract: Understanding how individual traits influence team performance is valuable, but these traits are not always directly observable. Prior research has in

Predictive Geometry of Hidden Trajectories in Transformers

arXiv:2609.37717v1 Announce Type: cross Abstract: Decoder-only transformers are trained only through a terminal next-token prediction loss, yet this loss constrains every intermediate hidden state thr

Tutorials

Pretraining Latent Information Feedback Transformers with Teacher Supervision

arXiv:2609.38149v1 Announce Type: new Abstract: Transformer language models (LMs) are feed-forward: deep-layer representations are never fed back to shallower layers, and the only pathway for informat

tutorialsarxiv-cs-cl

ProgressCompass: Embodied Progress Reward Models Are Lost Without the Right Context

arXiv:2609.36684v1 Announce Type: new Abstract: Embodied agents now take on ever longer tasks. For long tasks, knowing only whether a task finally succeeds or fails says little; the steps along the wa

Pruning for Efficiency, Paying in Fairness: Demographic Disparities in Pruned Speech-LLMs

arXiv:2609.38106v1 Announce Type: cross Abstract: Speech-LLMs are expensive to run, making compression important for real-world deployment. However, compressed models are usually selected using aggreg

QuantMLA: Function-Aligned Dual-Path Quantization for Low-Bit MLA KV Caching

arXiv:2609.36760v1 Announce Type: cross Abstract: Multi-Head Latent Attention (MLA) enables expressive multi-head attention with compact caches for its content and decoupled RoPE paths, yet cache memo

Question-Specific Knowledge Graphs for Efficient Visual Reasoning

arXiv:2609.35942v1 Announce Type: new Abstract: Recent work in visual question answering has shown that vision-language models can exhibit strong reasoning capabilities by translating visual inputs in

RAEGNet: Relation-Aware Evidence Graph Network for Harm-Aware Multimodal Fake News Detection

arXiv:2609.36902v1 Announce Type: new Abstract: Existing multimodal fake news detection methods often introduce external information to assist detection. However, most of them rely on entity-level ret

RAWR: Reward Assignment Without Rollouts in Verifiable Domains

arXiv:2603.17815v2 Announce Type: replace Abstract: Understanding and evaluating multi-step reasoning in LLMs at the level of individual steps remains a key challenge. Process reward models (PRMs) pro

Reader Proficiency Shapes Layer-wise Surprisal Profiles

arXiv:2609.37688v1 Announce Type: new Abstract: Reading behaviour varies not only with linguistic input, but also with reader proficiency. In this study, we investigate whether the layer-wise relation

Local Ai

Reconstructing the Vocal Tract with Differentiable Acoustic Simulation

arXiv:2609.36737v1 Announce Type: cross Abstract: The vocal tract is the region of the human body responsible for filtering one's voice to create speech. In this paper, we present a differentiable and

local-aiarxiv-cs-cl

Relevance Is Not Sufficient Evidence: Detecting Evidence Gaps Before Generation in RAG

arXiv:2609.37469v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) grounds large language models in external sources, but retrieved passages often name the right entities without pro

Reliable but Design-Sensitive: Instrument Uncertainty in LLM Annotation

arXiv:2609.35824v1 Announce Type: new Abstract: Large language models (LLMs) can give reliable labels under one setup yet change those labels when researchers make other reasonable design choices. We

Repetition, Not Length: Isolating the Counting Failure in Neural Text-to-Speech

arXiv:2609.36974v1 Announce Type: new Abstract: Text-to-speech models loop, truncate and lose count on text that repeats a phrase many times. We show that repetition itself is what breaks them, not th

Replay the Curvature: Accurate and Scalable NVFP4 Quantization for Large Language Model Inference

arXiv:2609.36654v1 Announce Type: cross Abstract: Large language models make weight storage and memory traffic major inference costs, motivating low-precision formats that represent each weight with o

Resolving the Missing Financial Data Crisis: A Generative AI Pipeline for SEC 10-K Extraction

arXiv:2609.35864v1 Announce Type: new Abstract: SEC 10-K filings contain substantial financial information that is not consistently captured in structured datasets, creating a missing-data problem aff

Applications

Rethinking Multimodal Fake News Detection in the Generative AI Era

arXiv:2609.36850v1 Announce Type: new Abstract: Generative content is increasingly entering the production and dissemination of news, transforming fake news from manually fabricated or simply manipula

applicationsarxiv-cs-cl

Retrieval Capacity of Self-Attention Under Competition

arXiv:2609.37879v1 Announce Type: new Abstract: How many tokens from its context does a language model actually use, and what determines that number? We study this question through self-attention. Wit

Retrieval Sensitivity to Identity Signals in Queries

arXiv:2609.36534v1 Announce Type: new Abstract: Dense retrievers decide which documents reach users and the language models that use them, yet they are typically evaluated with neutral queries. We ask

RunyaNER: Auxiliary Language Selection for Runyankore NER

arXiv:2609.37543v1 Announce Type: new Abstract: Cross-lingual zero-shot transfer and multilingual fine-tuning are promising approaches for NLP tasks such as Named Entity Recognition (NER) in low-resou

SEED: Self-Speculative Decoding via Implicit Encoder-Decoder

arXiv:2609.36590v1 Announce Type: new Abstract: Self-speculative decoding accelerates large language model (LLM) inference by drafting tokens from the target model itself, but faces a sharp tradeoff b

Seeing What Should Be Heard: Diagnosing and Repairing Cross-Modal Shortcuts in Omni-Modal LLMs

arXiv:2609.36798v1 Announce Type: cross Abstract: Omni-modal large language models (LLMs) are expected to answer a question using the modality it explicitly refers to. However, existing training parad

Selecting What Matters: Semantic Compression-Guided Selective Pooling for Long-Context Embeddings

arXiv:2609.37782v1 Announce Type: new Abstract: Large language models (LLMs) have shown strong potential as training-free text encoders for long-context embeddings. Existing approaches primarily impro

Applications

SemOPT: Fixing Semantic Errors in LLM-based Optimization Modeling via Reward-Guided Search

arXiv:2609.37361v1 Announce Type: new Abstract: Operations research supports decision-making in domains such as energy, economics, and healthcare. Solving operations research problems typically begins

applicationsarxiv-cs-cl

SiDiaC-v.2.0: Sinhala Diachronic Corpus Version 2.0

arXiv:2603.10861v2 Announce Type: replace Abstract: SiDiaC-v.2.0 is the largest comprehensive Sinhala Diachronic Corpus to date, covering a period from 1800 CE to 1955 CE in terms of publication dates