Search: “arxiv-cs-cl”

NestedKV: Nested Memory Routing for Long-Context KV Cache Compression

arXiv:2605.26678v1 Announce Type: new Abstract: Long-context language models are limited by the memory footprint of the key-value (KV) cache. Existing training-free KV compression methods usually rank

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Not All Tokens Matter Equally: Dynamic In-context Vector Distillation with Decisive-Token Supervision for Long-form Medical Report Generation

DGX agent

arXiv:2605.27194v1 Announce Type: new Abstract: Distilling demonstration effects into hidden-space interventions offers a lightweight alternative to full finetuning. However, existing multimodal varia

researcharxiv-cs-cl

27 May 2026

NSF-SciFy: Mining the NSF Awards Database for Scientific Claims

arXiv:2503.08600v3 Announce Type: replace Abstract: We introduce NSF-SciFy, a comprehensive dataset of scientific claims and investigation proposals extracted from National Science Foundation award ab

OmniInteract: Benchmarking Real-World Streaming Interaction for Real-Time Omnimodal Assistants

arXiv:2605.26485v1 Announce Type: cross Abstract: We introduce OmniInteract, a streaming benchmark for real-time omnimodal large language models evaluated through native online inference over audio-vi

On the Hidden Costs of Counterfactual Knowledge Training in LLM Unlearning

arXiv:2605.27083v1 Announce Type: new Abstract: Counterfactual tuning (CFT) has emerged as a promising paradigm for Large Language Model (LLM) unlearning by training models to generate alternative fic

On the Sensitivity of Instruction-tuned LLMs to Harmful Sentences in Long Inputs

arXiv:2510.05864v2 Announce Type: replace Abstract: Large language models (LLMs) increasingly operate on long inputs, yet their behavior when harmful sentences are sparsely embedded within such inputs

Tutorials

Optimising Factual Consistency in Summarisation via Preference Learning from Multiple Imperfect Metrics

arXiv:2605.26840v1 Announce Type: new Abstract: Reinforcement learning with evaluation metrics as rewards is widely used to enhance specific capabilities of language models. However, for tasks such as

tutorialsarxiv-cs-cl

PashtoTTS-Bench: automated screening for low-resource non-Latin-script text-to-speech

arXiv:2605.26978v1 Announce Type: new Abstract: Text-to-speech (TTS) evaluation for low-resource non-Latin-script languages can fail when it relies on a single ASR round-trip word error rate (WER). A

PersianMedQA: Evaluating Large Language Models on a Persian-English Bilingual Medical Question Answering Benchmark

arXiv:2506.00250v4 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved remarkable performance on a wide range of Natural Language Processing (NLP) benchmarks, often surpassing

PersLitEval: Fine-grained Benchmark and Evaluation of LLMs on Persian Literature Questions

arXiv:2605.27015v1 Announce Type: new Abstract: Despite impressive multilingual capabilities, large language models (LLMs) remain poorly evaluated on literary knowledge in non-English languages. We in

PinPoint: Prompting with Informative Interior Points

arXiv:2605.26689v1 Announce Type: cross Abstract: Modern referring image segmentation pipelines couple a vision-language model (VLM) for grounding with a promptable segmenter such as the Segment Anyth

Pop-Up Distractions Reveal Bag-of-Events Behavior in Video Large Language Models

arXiv:2605.27101v1 Announce Type: cross Abstract: A key capability for video understanding is reliably linking subjects to events across time, yet whether Video Large Language Models (VideoLLMs) actua

PRISM: A Multi-Dimensional Benchmark for Evaluating LLM Peer Reviewers

arXiv:2605.26730v1 Announce Type: new Abstract: The rapid growth in submissions to machine learning venues has strained the scientific peer-review system and intensified interest in LLM-based automate

Probing Cultural Awareness in LLMs: A Case Study of Cross-Culture Aesthetic Stylistics

arXiv:2605.27296v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed in diverse cultural contexts, yet their ability to master aesthetic stylistics, i.e., the strateg

Probing Minimalist Phase Structure in LLMs: What Universal Dependencies Cannot Represent

arXiv:2605.26431v1 Announce Type: new Abstract: Structural probes train on Universal Dependencies (UD), which does not encode formal-syntactic abstractions such as phase boundaries or phase-internal c

Probing the Knowledge Boundary: An Interactive Agentic Framework for Deep Knowledge Extraction

arXiv:2602.00959v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) can be seen as compressed knowledge bases, but it remains unclear what knowledge they truly contain and how far t

Prompt Injection Detection is Regime-Dependent: A Deployment-Aware Evaluation with Interpretable Structural Signals

arXiv:2605.26999v1 Announce Type: new Abstract: Prompt injection poses a critical threat to the safe deployment of large language models, yet existing detection approaches are typically evaluated unde

Psychological Constructs in Shared Semantic Space

arXiv:2605.26801v1 Announce Type: new Abstract: Psychological constructs are often measured in separate instruments, datasets, and research traditions, which makes direct comparison difficult. This pa

QAM-W: Joint 2D Codebook Quantization for LLM Weights via Hadamard Rotation and Activation-Aware Scaling

arXiv:2605.26339v1 Announce Type: cross Abstract: Scalar post-training quantizers discard pairwise coordinate structure within weight rows. We introduce QAM-W (Quadrature Amplitude Modulation for Weig

Quadratic Term Correction on Heaps' Law

arXiv:2511.14683v2 Announce Type: replace Abstract: Heaps' or Herdan's law characterizes the word-type vs. word-token relation by a power-law function, which is concave in linear-linear scale but a st

Quality Without Usefulness: LLM-Generated XAI Narratives as Trust Heuristics Rather Than Decision Aids

arXiv:2605.26770v1 Announce Type: new Abstract: Prior work shows that Large Language Models (LLMs) can transform Explainable AI (XAI) outputs into Natural Language Explanations (NLEs) that score highl

Real Images, Worse Judgments: Evaluating Vision-Language Models on Concreteness and Imagery

arXiv:2605.27315v1 Announce Type: new Abstract: Visual inputs are often assumed to improve language understanding in multimodal models. We examine this assumption by asking whether vision-language mod

Representation-Aware Unlearning via Activation Signatures: From Suppression to Entity-Signature Erasure

arXiv:2601.10566v5 Announce Type: replace Abstract: Entity-level unlearning is usually evaluated by what a model says: whether it stops naming the target, refuses a query, or shifts a Truth Ratio dist

Rethinking the Multilingual Reasoning Gap with Layer Swap

arXiv:2605.26735v1 Announce Type: new Abstract: Recent reasoning Large Language Models produce a chain-of-thought (CoT) predominantly in English, even when prompted in non-English languages. Prior wor

RICE-PO: Turning Retrieval Interactions into Credit Signals for Reasoning Agents

arXiv:2605.26352v1 Announce Type: new Abstract: Retrieval is increasingly moving from one-shot matching toward interactive reasoning, where language agents iteratively inspect evidence, reformulate qu

Self-Ensembling Vision-Language Models for Chart Data Extraction

arXiv:2605.27298v1 Announce Type: new Abstract: Charts effectively convey quantitative information, but the underlying data are often locked in image form, hindering reuse and analysis. Manually digit

Self-Verified Distillation: Your Language Model Is Secretly Its Own Synthetic Data Pipeline

arXiv:2605.26132v1 Announce Type: new Abstract: Can post-trained large language models (LLMs) further improve themselves using only unlabeled prompts, without external teachers or feedback from tools?

Applications

Semantic Gradients Interactions in SSD: A Case Study in Racial Identity and Hate Speech

arXiv:2605.27322v1 Announce Type: new Abstract: We introduce interaction SSD, an extension of Supervised Semantic Differential that models how semantic meaning varies across moderators such as groups,

applicationsarxiv-cs-cl

Separating Semantic Competition from Context Length in RAG Reading

arXiv:2605.27294v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) systems can respond incorrectly even when the correct passage was retrieved. The model must still read the retrieve

Tutorials

Share More, Search Less: Collaborative Parallel Thinking for Efficient Test-Time Scaling

arXiv:2605.27030v1 Announce Type: new Abstract: Test-Time Scaling (TTS) enhances the reasoning capabilities of large language models by allocating additional inference compute to explore the solution

tutorialsarxiv-cs-cl

Shopping Companion: A Memory-Augmented LLM Agent for Real-World E-Commerce Tasks

arXiv:2603.14864v2 Announce Type: replace Abstract: In e-commerce, LLM agents show promise for shopping tasks such as recommendations, budget management, and bundle deals, where accurately capturing u

Slide Deck Q&A Quality Assurance App: A Multi-Stage Pipeline for Pedagogical Question Generation

arXiv:2605.26428v1 Announce Type: new Abstract: Generating high-quality, pedagogically useful questions from lecture slide decks is difficult because important instructional content is distributed acr

SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learning

arXiv:2603.28730v2 Announce Type: replace-cross Abstract: Vision-language models (VLMs) have shown impressive capabilities across diverse tasks, motivating efforts to leverage these models to supervis

SONAR-LLM: Autoregressive Transformer that Thinks in Sentence Embeddings and Speaks in Tokens

arXiv:2508.05305v2 Announce Type: replace Abstract: The recently proposed Large Concept Model (LCM) generates text by predicting a sequence of sentence-level embeddings and training with either mean-s

SPEAR: Code-Augmented Agentic Prompt Optimization

arXiv:2605.26275v1 Announce Type: new Abstract: Automatic prompt engineering (APE) rewrites prompts to improve downstream task performance, but existing APE loops treat the optimizer itself as a fixed

Stylistic Evolution and LLM Neutrality in Singlish Language

arXiv:2601.06580v2 Announce Type: replace Abstract: Singlish is a creole rooted in Singapore's multilingual environment that continues to evolve alongside social and technological change. We examine d

SWE-Adept: An LLM-Based Agentic Framework for Deep Codebase Analysis and Structured Issue Resolution

arXiv:2603.01327v2 Announce Type: replace-cross Abstract: Large language models (LLMs) exhibit strong performance on self-contained programming tasks. However, they still struggle with repository-leve

Telenor Nordics Customer Service self-help corpus

arXiv:2605.26891v1 Announce Type: new Abstract: This paper presents a multilingual customer service self-help corpus comprising 1,122 manually validated documents in Finnish, Danish, Norwegian, and Sw

Temporal Simultaneity Predicts Annotation Quality in Sentiment Corpora

arXiv:2605.27239v1 Announce Type: new Abstract: Annotation quality is difficult to sustain when campaigns span weeks or months with small annotator pools. We present a Setswana sentiment dataset of 3,

The Coverage Illusion: From Pre-retrieval Routing Failure to Post-retrieval Cascades in a Production RAG System

arXiv:2605.27220v1 Announce Type: new Abstract: In modern RAG pipelines, query augmentation methods such as HyDE and query expansion are applied to every query, resulting in substantial LLM inference

The Daily Dose: Workflow-Integrated Large Language Model Automation for Clinical Summarization and Trial Identification in Radiation Oncology

arXiv:2605.26346v1 Announce Type: new Abstract: Objective: To describe the design and early clinical evaluation of The Daily Dose (TDD), an LLM-driven, automated clinical summarization and clinical-tr

The Need for an External Observer Formalizing the Sufficiency Gap: A Mathematical Extension of Mixture Identifiability and Contextual Grounding in Sequence Models

arXiv:2605.26711v1 Announce Type: new Abstract: We construct a binary mixed-regime process with one deterministic textual regime and one random regime governed by an unobserved latent state. Even an i

To model human linguistic prediction, make LLMs less superhuman

arXiv:2510.05141v2 Announce Type: replace Abstract: When we read, we make predictions about upcoming words; these predictions influence our reading behavior. The success of large language models (LLMs

Tutorials

Towards Just-in-Time Adaptive Feedback: Enhancing Student Learning via Knowledge-Grounded LLM

arXiv:2605.26405v1 Announce Type: new Abstract: Educational interventions are effective tools for enhancing student learning. While Large Language Models (LLMs) allow for generating adaptive feedback

tutorialsarxiv-cs-cl

Applications

UltraCUA: A Foundation Model for Computer Use Agents with Hybrid Action

arXiv:2510.17790v3 Announce Type: replace-cross Abstract: Computer-use agents face a fundamental limitation. They rely exclusively on primitive GUI actions (click, type, scroll), creating brittle exec

applicationsarxiv-cs-cl

Uncertainty-Aware Budget Allocation for Adaptive Test-Time Reasoning

arXiv:2605.26849v1 Announce Type: new Abstract: Sampling multiple responses improves language model reasoning, but uniform compute allocation is inefficient: easy questions are over-sampled while hard

Vectors Are Not Neutral: Sensitive-Information Inference from Exported LLM Representations in Summarization

arXiv:2605.26433v1 Announce Type: new Abstract: Large language model (LLM) summarization systems may pass compact vector representations of private inputs to downstream retrieval, monitoring, audit, o

VERA-V: Variational Inference Framework for Jailbreaking Vision-Language Models

arXiv:2510.17759v2 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) extend large language models with visual reasoning, but their multimodal design also introduces new, underexplor