Search: “arxiv-cs-cl”

A Generative Grammar Underlying the Voynich Manuscript, the Pastiche Hypothesis: Evidence from Large Language Models

arXiv:2609.20835v1 Announce Type: new Abstract: Background: The Voynich Manuscript is a fifteenth-century codex written in an unknown script whose content remains undeciphered. Previous studies sugges

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

Research

Abstention and Noise Filtering: Two Missing Primitives of Softmax Attention

DGX agent

arXiv:2609.22005v1 Announce Type: cross Abstract: Gating the value pathway of attention reportedly improves language model pretraining, and prior studies disagree on why. We argue and provide experime

researcharxiv-cs-cl

21 Sept 2026

An Interpretable Memory Decision Controller for LLM Agents Based on Three-Signal Complementarity: Decoupling Confidence and Consistency

arXiv:2609.22043v1 Announce Type: new Abstract: Memory systems for large language models have focused predominantly on efficient retrieval, whereas the decision of whether retrieved memories should be

Analysing the Linearity of Linguistic Relations in Language Model Embedding Spaces

arXiv:2609.21655v1 Announce Type: new Abstract: We propose a framework to analyse how strongly different linguistic relations are linearly encoded in language model embedding spaces. We formalise line

ArenaFlow: From Trajectory Ranking to Hierarchical Credit Propagation for Open-Ended Agent RL

arXiv:2609.21378v1 Announce Type: new Abstract: Reinforcement learning has substantially improved large language model (LLM) agents in verifiable domains, but remains difficult to apply to open-ended

Benchmarking Gender Bias in Machine Translation Evaluation Metrics across Occupations

arXiv:2609.21490v1 Announce Type: new Abstract: Gender bias remains a persistent concern in machine translation (MT), affecting both generated translations and their automatic evaluation. When a sourc

Beyond Atomic Tokens: Factorizing Syllables for Language Model Pretraining

arXiv:2609.21362v1 Announce Type: new Abstract: Conventional tokenizers represent text as characters or statistically derived subwords, overlooking the internal phonological structure of syllables and

Beyond Reference-Based Evaluation: Reward Models for Meta-Evaluation of Grammatical Error Correction

arXiv:2609.21231v1 Announce Type: new Abstract: Reference-based metrics for Grammatical Error Correction (GEC) such as M^2 and ERRANT assume that the reference set enumerates all valid edits, and ther

Beyond WER: Entity and Disfluency Recall in Accented Conversational ASR

arXiv:2609.20828v1 Announce Type: new Abstract: ASR systems optimised for Word Error Rate (WER) often miss named entities and filled pauses in accented conversational English, both critical for langua

Boosting Deepresearch and LongContext Ability with Self-Generated Deepresearch Rollouts Traces

arXiv:2609.20844v1 Announce Type: new Abstract: Deepresearch (DR) agents interact with real-world web environments through multi-turn search and visit, causing their contexts to grow rapidly over time

CASCADE Against Jailbreaks: Combination Across Stages with Controlled Attack-Defense Evaluation

arXiv:2609.21793v1 Announce Type: cross Abstract: Defenses against jailbreak attacks on Large Language Models (LLMs) operate at different pipeline stages, such as input modification or output guard, b

Applications

COAL-SQL: Coverage-Guided Augmentation and Failure-Driven Learning for Text-to-SQL Post-Training

arXiv:2609.20842v1 Announce Type: new Abstract: Text-to-SQL translates natural-language questions into executable SQL queries, but open-source large language models still require task-specific post-tr

applicationsarxiv-cs-cl

CoLearn: An Agentic Tutor that Learns its Learner in a Human--AI Co-Learning Loop

arXiv:2609.21154v1 Announce Type: new Abstract: Good tutoring adapts to the individual: it tracks what a learner knows, notices why they go wrong, and asks the next question that will help most. Most

Applications

Configurable Multi-Stage Vision Pipeline for Crop Disease and Pest Diagnosis

arXiv:2609.21651v1 Announce Type: cross Abstract: Farmer.Chat is Digital Green's farm advisory service for smallholder farmers. When something looks wrong with a crop, the farmer takes a photograph an

applicationsarxiv-cs-cl

Conformal Privacy Auditing: Calibrated Re-identification Attacks with Statistical Guarantees

arXiv:2609.21340v1 Announce Type: cross Abstract: Empirical identity leakage from released text is increasingly driven by attackers that combine large language models (LLMs) with auxiliary knowledge t

Cross-Lingual Parkinson's Disease Severity Assessment Using Pre-trained Speech Embeddings: A Multi-Class Evaluation

arXiv:2609.20875v1 Announce Type: cross Abstract: Parkinson's disease (PD) often manifests through speech impairments, facilitating accessible, non-invasive, and cost-effective severity assessment for

Cross-sector generalization of accident-process role classification in occupational accident narratives

arXiv:2609.22081v1 Announce Type: new Abstract: Occupational accident narratives contain valuable information about work situations, unfavourable conditions, accident events, and their consequences. A

Local Ai

Cultural Alignment in Large Language Models Using Soft Prompt Tuning

arXiv:2503.16094v2 Announce Type: replace Abstract: Large Language Model (LLM) alignment is commonly achieved through supervised fine-tuning or reinforcement learning, both of which require labeled or

local-aiarxiv-cs-cl

Curriculum-Based Noise Adaptation for Phoneme-to-Text Reconstruction in Visual Speech Recognition

arXiv:2609.20839v1 Announce Type: new Abstract: Phoneme-centric visual speech recognition reconstructs sentences from intermediate phoneme predictions, making overall recognition performance highly de

Do small language models know what they don't know?

arXiv:2609.20824v1 Announce Type: new Abstract: We explore whether entropy-based confidence signals can be leveraged to improve the accuracy of Small Language Models (SLMs) with fewer than 3 billion p

Dynamic Lagging using Stable-Prefix Training for Simultaneous Translation

arXiv:2609.05799v2 Announce Type: replace Abstract: In streaming simultaneous speech translation, the speech translation system is trained to learn a read-write policy that alternates between consumin

Enhancing Audio Reasoning via Semantic Summary Prediction

arXiv:2609.20849v1 Announce Type: new Abstract: Large Audio Language Models (LALMs) perform well on complex question answering but often show a reasoning gap, where explicit Chain-of-Thought (CoT) red

Evaluating In-Context Learning and Retrieval Strategies for Devanagari Post-OCR Correction

arXiv:2609.21595v1 Announce Type: new Abstract: In-context learning using Large Language Models (LLMs) offers a compelling path to training-free post-OCR correction, yet its effectiveness for Devanaga

Explainable Multimodal Aspect-Based Sentiment Analysis with Dependency-guided Large Language Model

arXiv:2601.06848v2 Announce Type: replace Abstract: Multimodal aspect-based sentiment analysis (MABSA) aims to identify aspect-level sentiments by jointly modeling textual and visual information, whic

FairLMs: A Turnkey Library for Fairness in Language Models

arXiv:2609.21296v1 Announce Type: cross Abstract: Fairness research on language models involves measuring bias, applying mitigation methods, and examining the evidence on which an evaluation rests. Ex

From Discharge Notes to Patient Understanding: Persona-Grounded, Open-Ended Simulation of LLMs as Discharge Educators

arXiv:2609.20827v1 Announce Type: new Abstract: Hospital discharge education is an interactive teaching task: a clinician adapts a discharge plan to a patient's literacy, recall, and personality. Exis

From Generation to Detection: Exploration of Discourse Driven Scenario based LLM Generated Fake News

arXiv:2609.20838v1 Announce Type: new Abstract: In this study, we examine how modern LLMs generate and detect fake news under controlled settings across four manipulation scenarios. These are open-end

Generative Artificial Intelligence Chatbots for Motivational Interviewing: A Scoping Review From System Design to Intervention Outcomes

arXiv:2609.20902v1 Announce Type: new Abstract: Motivational interviewing (MI) is a collaborative approach to elicit autonomous motivation for health behavior change. Generative AI (GenAI) offers new

HERMES: Contrast-Aware Knowledge Graph Reasoning from Clinical Notes for Patient Outcome Prediction

arXiv:2609.20825v1 Announce Type: new Abstract: Clinical predictive models often rely on structured Electronic Health Record data, such as time-series and procedure codes. While recent approaches have

Hierarchical attention interpretation: an interpretable speech-level transformer for bi-modal depression detection

arXiv:2309.13476v3 Announce Type: replace Abstract: Depression is a common mental disorder. Automatic depression detection tools using speech, enabled by machine learning, help early screening of depr

How Many Humans Is a Judge Panel Worth?

arXiv:2609.21277v1 Announce Type: new Abstract: How many human judgments does a panel of language models represent? The answer depends on what is matched. We audit categorical judge panels against emp

I'll Keep an Ear Out: Teaching AudioLLMs Proactive Audio Assistance

arXiv:2609.21183v1 Announce Type: cross Abstract: Audio large language models (AudioLLMs) operate reactively, responding only when queried. We introduce proactive audio assistance, where an AudioLLM m

Measuring Digital Labour Market Transitions with a Digital Semantic Score: An AI-Based Methodology Applied to the Dutch Labour Market

arXiv:2608.24222v2 Announce Type: replace Abstract: The digital transformation of the Dutch labour market is reshaping occupational language, career pathways, and job-related skills. Addressing these

MemoryArena: Benchmarking Agent Memory in Interdependent Multi-Session Agentic Tasks

arXiv:2602.16313v2 Announce Type: replace Abstract: Existing evaluations of agents with memory typically assess memorization and action in isolation. One class of benchmarks evaluates memorization by

MIRAGE: Multi-Perspective Creative Language Model Reasoning with Reinforcement Learning Guidance

arXiv:2609.21554v1 Announce Type: new Abstract: Recent advances in Large Language Models (LLMs) have revolutionized artificial intelligence and how human interact with AIs. Despite impressive advancem

MME-Safety: A Fine-grained Benchmark for Safety Evaluation of MLLMs

arXiv:2609.20850v1 Announce Type: new Abstract: While Multimodal Large Language Models (MLLMs) show remarkable advancements, their cross-modal capabilities introduce complex vulnerabilities that easil

Moral Entropy: Auditing Bias and Uncertainty in Moral Judgment

arXiv:2609.21992v1 Announce Type: new Abstract: Most work in computational ethics treats annotator disagreement on moral content as noise to be voted away, collapsed into majority vote or the more per

mu^2-Bench: A Multilingual Machine Unlearning Benchmark

arXiv:2609.20945v1 Announce Type: new Abstract: Undesired information such as harmful content and private data propagates through Multilingual Large Language Models (LLMs) via direct training and indi

Not All Irregularity Is Equal: Causally Isolating a Rare Failure Mode in Japanese Morphological Inflection

arXiv:2609.21179v1 Announce Type: new Abstract: Neural morphological generation systems often achieve high aggregate accuracy on benchmark datasets, yet such performance can conceal systematic errors

Omni Demand Understanding: A Benchmark for Contextual User-Intent Inference in Multimodal Interaction

arXiv:2609.21392v1 Announce Type: new Abstract: Natural audio-visual interaction is emerging as an important interface for AI assistants, allowing users to communicate through speech and vision rather

Per-Aetiology Contrastive Severity Embeddings with Phonological Pseudo-Labelling for Multilingual Dysarthric Speech

arXiv:2609.21789v1 Announce Type: new Abstract: Most multilingual dysarthria-severity systems either train on a single aetiology-language pair or pool heterogeneous aetiologies into one label space. W

PhysioBench: A Unified Benchmark for Physiological Signal Question Answering

arXiv:2609.20836v1 Announce Type: new Abstract: Physiological signals support diverse clinical and monitoring tasks, yet existing physiological signal foundation models typically require task-specific

Predictable Failure in Multi-Hop Retrieval: Score-Distributional Confidence Scoring and Abstention

arXiv:2609.22056v1 Announce Type: cross Abstract: Multi-hop retrieval failures are not uniformly distributed across queries: they cluster in structurally predictable subpopulations. We prove two resul

Prediction Dynamics in Depth-Recurrent Language Models

arXiv:2609.21383v1 Announce Type: new Abstract: Depth-recurrent language models refine predictions through repeated latent updates. Why can intermediate answers agree with the endpoint while their sco

PRISM-BN: A Controlled Corpus and Benchmark for Text-to-Parameterized Bayesian Network Extraction

arXiv:2609.21673v1 Announce Type: new Abstract: Probabilistic Graphical Models (PGMs), especially Bayesian Networks (BNs), expose directed structure and probabilistic parameters, making them natural s

QuranicMMLU: A Cognitively-Aware Benchmark for Evaluating Generative AI Solutions on Quranic Linguistic Knowledge

arXiv:2609.22038v1 Announce Type: new Abstract: We introduce QuranicMMLU, a benchmark for evaluating generative AI on Quranic Arabic across multiple dimensions of linguistic complexity. Existing Quran

Reading Anxiety or Reading the Label? Comparing Fine-Tuned and Frontier Models for Anxiety Detection on Social Media

arXiv:2609.20847v1 Announce Type: new Abstract: Anxiety is among the most common mental health conditions, and people often write about it online well before seeking clinical help. Practitioners build

Reading Less While Writing: A Closed-Form Bandwidth Dial for Streaming Multimodal Decoders

arXiv:2609.20845v1 Announce Type: new Abstract: A decoder that turns video or audio into text conventionally consumes the entire input before emitting a word. Offline this is merely more than the task