Search: “arxiv-cs-cl”

RecreationWorld: Scalable and Verifiable Environments for Hybrid Computer-Use Agents

arXiv:2609.22000v1 Announce Type: new Abstract: Computer-use agents (CUAs) have advanced along two separate lines: graphical interaction and software development through code and the command line. Rea

Human

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Safety

Recursive Language Models Generalize Out of Domain

DGX agent

arXiv:2609.20831v1 Announce Type: new Abstract: We study when limiting what a language model can see improves learning. We compare standard CoT, the more general learner that reads the full trace, wit

safetyarxiv-cs-cl

21 Sept 2026

Rethinking Human-Aligned Evaluation: An Analysis of Semantic Metrics Beyond WER

arXiv:2609.21663v1 Announce Type: new Abstract: Word Error Rate (WER), the most commonly used metric for Automatic Speech Recognition (ASR), treats every lexical deviation from the reference as equall

Reusing Latent Speech Representations for Query-Conditioned Topic Localization in Transcripts

arXiv:2609.21844v1 Announce Type: new Abstract: Long transcripts are costly inputs for downstream NLP systems and often contain irrelevant context. We study query-conditioned topic localization: predi

Reviser: Revision-Capable Text Generation via Autoregressive Cursor Actions

arXiv:2609.20830v1 Announce Type: new Abstract: Revision-capable generation is appealing because it can insert or revise earlier content, but many non-autoregressive and edit-based approaches obtain t

Rewarding Efficient Reasoning Improves Abstention on Underspecified Tasks in Reasoning Models

arXiv:2609.20846v1 Announce Type: new Abstract: While modern large reasoning models (LRMs) excel at providing correct answers in many tasks, we provide additional evidence for the observation that the

RheoSampling: Resolving the One-Hot Dilemma in Stochastic Dynamic-Tree Speculative Decoding

arXiv:2609.21827v1 Announce Type: new Abstract: Speculative decoding accelerates LLM inference by drafting multiple tokens in parallel, with tree-based methods further improving efficiency through hie

SAGE: Schema-Guided LLMs for Grant Review

arXiv:2609.20829v1 Announce Type: new Abstract: Grant reviewers must apply detailed criteria to application forms, budgets, and supporting documents while producing assessments that colleagues can ins

Scaling Forced Alignment to End-User Devices

arXiv:2609.21145v1 Announce Type: new Abstract: The Viterbi algorithm has been previously used to perform forced alignment of audio to text to mine training data from online resources. However, many e

Semantic Calibration Prevails Where Token Confidence Fails: Benchmarking Long-Form Scientific QA

arXiv:2602.00279v2 Announce Type: replace Abstract: Reliable uncertainty quantification (UQ) is essential for safe deployment of large language models (LLMs) in scientific question answering, where lo

Souper-Model: How Simple Arithmetic Unlocks State-of-the-Art LLM Performance

arXiv:2511.13254v2 Announce Type: replace Abstract: Large Language Models (LLMs) have displayed remarkable capabilities across diverse domains, but their training remains resource- and time-intensive,

SteganoBackdoor: Evading Data-Poisoning Defenses via Steganographic Backdoors

arXiv:2511.14301v4 Announce Type: replace-cross Abstract: Transformer-based models are highly susceptible to backdoor attacks via supervised fine-tuning (SFT). To red-team existing data-poisoning defe

TALON: A Temporally Aware Longitudinal Framework for Radiology Report Generation

arXiv:2609.20826v1 Announce Type: new Abstract: Current radiology report generation (RRG) models usually produce descriptive reports based on a single examination or only the most recent prior examina

TatBLiMP: A Benchmark of Linguistic Minimal Pairs for Tatar

arXiv:2609.20832v1 Announce Type: new Abstract: We introduce TatBLiMP, the first benchmark of linguistic minimal pairs for Tatar (tt, ISO 639-3 tat), a Qypchaq Turkic language written in Cyrillic. To

The Hidden Cost of Digits: Number Normalization and WER in ASR Systems

arXiv:2609.21084v1 Announce Type: cross Abstract: Modern automatic speech recognition (ASR) systems trained on extremely large datasets can produce transcripts with numbers written in Arabic numerals.

The Spoken Wikipedia Presentation Corpus

arXiv:2609.21676v1 Announce Type: cross Abstract: We present the Spoken Wikipedia Presentation Corpus, an extension of the Spoken Wikipedia Corpora featuring LLM-generated slide decks for multimodal A

Towards Secure Cloud-Native Computing: Unveiling Kubernetes Misconfigurations with Large Language Models

arXiv:2609.20834v1 Announce Type: new Abstract: In the rapidly evolving landscape of cloud-native computing, Organizations are increasingly adopting infrastructure models that emphasize scalability, f

Local Ai

Transsion's Speaker-Attributed Multilingual ASR System for the MLC-SLM 2026 Challenge

arXiv:2609.20833v1 Announce Type: new Abstract: This paper presents the Transsion Speech Team submission to Task 1 of the MLC-SLM 2026 Challenge, which focuses on speaker-attributed transcription for

local-aiarxiv-cs-cl

TrialAtlas: Multi-Agent Research Organization for Clinical Trial Design and Optimization

arXiv:2609.21859v1 Announce Type: new Abstract: Nearly 90% of drugs entering clinical development ultimately fail, despite billions of dollars in investment. Pharmaceutical companies therefore rely on

VISPATH: Visual-Intent-Guided Path Reasoning for Multimodal Knowledge Graph Question Answering

arXiv:2609.20843v1 Announce Type: new Abstract: Knowledge graph question answering (KGQA) enables models to answer natural-language questions through structured graph reasoning and has achieved substa

Voice-Light: A Full-Duplex Cascaded Voice Agent with Causal Turn-Taking and Speculative Generation

arXiv:2609.20995v1 Announce Type: cross Abstract: Natural spoken interaction requires more than streaming ASR, language generation, and speech synthesis: a system must react to overlap without canceli

VQ-Logits: Compressing the Output Bottleneck of Large Language Models via Vector Quantized Logits

arXiv:2505.10202v2 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved remarkable success but face significant computational and memory challenges, particularly due to their ex

When Better Turns Do Not Make Better Agents: Diagnosing the Gap Between Next-Turn Metrics and Workflow Success

arXiv:2609.21187v1 Announce Type: new Abstract: Agent models are frequently evaluated one decision at a time, where the model predicts the next action based on the gold interaction history, which is s

When Does Reasoning Help in Machine Translation? A Hierarchical Analysis of LRM Reasoning Traces

arXiv:2609.21247v1 Announce Type: new Abstract: Large Reasoning Models increasingly use intermediate traces for machine translation, but it remains unclear when such reasoning helps or hurts. We analy

A Cross-Lingual Acoustic Disease-Alignment Framework for Respiratory Health Assessment from Spontaneous Speech

arXiv:2609.19398v1 Announce Type: cross Abstract: Spontaneous speech offers a scalable, noninvasive signal for respiratory health assessment, yet interpretable models that generalize across languages

Agents

A frontend-backend architecture for tool calls in full-duplex speech models

arXiv:2609.19334v1 Announce Type: new Abstract: Full-duplex speech-to-speech (S2S) models provide natural, low-latency conversational interaction and would benefit from the ability to use external too

agentsarxiv-cs-cl

A Large-Scale Vision-Language Dataset Derived from Open Scientific Literature to Advance Biomedical Generalist AI

arXiv:2503.22727v3 Announce Type: replace Abstract: Despite the excitement behind biomedical artificial intelligence (AI), access to high-quality, diverse, and large-scale data - the foundation for mo

A Phonemically Comprehensive, ASCII-Only Romanization Scheme for Thai and Lao: Systematic Cross-Lingual Correspondence and Chinese-User-Friendly Design

arXiv:2609.19736v1 Announce Type: new Abstract: This paper proposes a phonemically comprehensive, ASCII-only romanization scheme for Thai and Lao, treating the two closely related languages as a unifi

Advantage Scale Calibration Imbalance in Group-Relative Optimization under Low-Variance Rewards: Diagnosis and Bounded Recovery

arXiv:2609.19164v1 Announce Type: new Abstract: In verifier-style RLVR, group-relative optimization often treats advantage scale as an implementation detail. This paper separates two low-variance case

An Analysis of Training-Free Self-Reported Confidence in Language Models

arXiv:2609.20541v1 Announce Type: new Abstract: Large language models can report a numerical confidence together with generated content, but it is unclear whether this report is more than calibrated r

Are Finer Citations Always Better? Rethinking Granularity for Attributed Generation

arXiv:2604.01432v3 Announce Type: replace Abstract: Citation granularity -- whether to cite individual sentences, paragraphs, or documents -- is a critical design choice in attributed generation. Whil

Automated Gradient-Driven Parameter Sharing for Low-Resource Multilingual Speech-to-Text Translation

arXiv:2603.25836v2 Announce Type: replace Abstract: In low-resource multilingual speech-to-text translation, uniform architectural sharing across languages frequently introduces representation conflic

Applications

Before the Arrest: Benchmarking LLMs on Criminal Profiling from Incomplete Evidence

arXiv:2609.19965v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly applied to legal and criminal justice tasks, yet existing work focuses almost exclusively on post-arrest s

applicationsarxiv-cs-cl

Before the Warning Comes Too Late: Incremental Phone-Scam Detection from Speech

arXiv:2609.20223v1 Announce Type: new Abstract: We study weakly supervised incremental telecom fraud detection from raw telephone audio, where training provides only conversation-level labels and pred

Benchmarking LLM Compliance with China AI Generated Content Regulations

arXiv:2609.19989v1 Announce Type: new Abstract: The widespread adoption of LLMs has led to escalating content compliance risks. Prior works have contributed to addressing these risks in the English co

BurnRiSc: Toward Non-Invasive Burnout Screening in Open Source from Public Repository Signals

arXiv:2609.19422v1 Announce Type: cross Abstract: Burnout is a chronic occupational syndrome, and open source is close to a worst case for it: maintainers absorb unbounded demand with no manager to re

Chain-of-Thought Entropy as a Reliability Signal: A Preregistered Reproduction

arXiv:2609.19606v1 Announce Type: new Abstract: This empirical study is an independent reproduction of the dissociation Zhao reported in 2026. The shape of a large language model's chain-of-thought en

Compass-v3: Scaling Domain-Specific LLMs for Multilingual E-Commerce in Southeast Asia

arXiv:2509.09121v2 Announce Type: replace Abstract: Large language models (LLMs) excel in general-domain applications, yet their performance often degrades in specialized tasks requiring domain-specif

CounselReflect: Opportunities and Challenges for Designing Tools to Support Self-Reflection on Mental Health and Well-Being Conversations with AI

arXiv:2603.29429v2 Announce Type: replace Abstract: AI is increasingly used for mental health and well-being support, creating an urgent need for safer engagement, while design, evaluation, and govern

Agents

CovR: Coverage-Aware Hardware Verification via Reasoning-Guided Reinforcement Learning

arXiv:2609.19189v1 Announce Type: cross Abstract: Design verification remains one of the most resource-intensive stages of hardware development, often consuming up to 70% of the total design effort. W

agentsarxiv-cs-cl

D-Quant: Driftable Entropy Coding for KV Cache Quantization

arXiv:2609.19880v1 Announce Type: new Abstract: The KV cache has become a major bottleneck in deploying LLMs, as its memory footprint grows linearly with sequence length and batch size, imposing subst

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

arXiv:2609.19969v1 Announce Type: new Abstract: The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost

Design of the IBM Granite 5.0 TurboCTC ASR Model

arXiv:2609.20104v1 Announce Type: new Abstract: We describe the architecture, training methodology and inference speedups of Granite 5.0 Turbo CTC, a 470 million parameter encoder-only model with an e

Dictionary-Constrained Grapheme-to-Phoneme for Unsegmented Languages from LLM-Annotated Data

arXiv:2609.19805v1 Announce Type: new Abstract: Grapheme-to-phoneme (G2P) conversion turns raw text into its phonemic form and is an essential part of both text-to-speech (TTS) and automatic speech re

dQwen3.5: Hybrid-Attention Diffusion Language Models

arXiv:2609.20751v1 Announce Type: new Abstract: Adapting a pretrained autoregressive (AR) model is a cost-efficient route to a diffusion language model (DLM). While nearly all such adaptations start f

Embedding Models Measure in Peculiar Ways

arXiv:2609.20821v1 Announce Type: new Abstract: Embedding spaces define notions of semantic similarity and distance. We study whether those embeddings reflect physical measurements of mass, distance,

Evaluating Communicative Success in Machine-Translated Conversation

arXiv:2609.19885v1 Announce Type: new Abstract: Interpreter agents built on machine translation (MT) increasingly mediate live conversation between people who do not share a language, yet we still eva

F^{2}DR: A Fine-Grained Full-Pipeline Reward Framework for DeepSearch Workflows

arXiv:2609.19827v1 Announce Type: new Abstract: With the widespread industrial deployment of Large Language Models (LLMs), DeepSearch has emerged as the dominant paradigm for resolving complex user qu