Search: “arxiv-cs-cl”

Looking under the Wrong Lamppost: On the Limitations of Automated Translation Quality Estimation

arXiv:2608.03577v1 Announce Type: new Abstract: Automation of Translation Quality Estimation (QE) has emerged as a widely discussed approach to managing translation quality at scale, and a growing num

Model Releases

LoopMTP: A looped transformer guided by latent multi-token prediction

DGX agent

arXiv:2608.03624v1 Announce Type: new Abstract: Looped transformers have emerged as a parameter-efficient alternative to scaling depth for strong reasoning. By reusing one stack of layers across T ite

model-releasesarxiv-cs-cl

5 Aug 2026

M-GATE: Multilingual Grammar, Accuracy in Translation, and Efficiency Benchmark for Large Language Models

arXiv:2608.03803v1 Announce Type: new Abstract: Multilingual language models are deployed across a hundred or more languages, yet most benchmarks test whether a model can perform a task _in_ a languag

Mapping the City Through the Lens of Language Models

arXiv:2608.02971v1 Announce Type: new Abstract: Language models often complete an underspecified reference to a city with unstated assumptions about urban size, form, infrastructure, environment, and

MIDI-LLM: Improving Text-to-MIDI Music Generation via Adapting Large Language Models

arXiv:2511.03942v2 Announce Type: replace-cross Abstract: We present MIDI-LLM, a recipe that improves multitrack text-to-MIDI generation via adapting Large Language Models (LLMs). MIDI-LLM expands an

MoEGen: Mixture-of-Experts for Instance-Adaptive LoRA Generation

arXiv:2608.03275v1 Announce Type: new Abstract: Parameter-efficient fine-tuning (PEFT) enables efficient adaptation of large language models, but existing MoE-based PEFT methods typically improve capa

On the Diversity of Analogy Making in Large Language Models

arXiv:2608.03233v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated remarkable potential for analogy making, a core cognitive capability that drives novelty and creativity.

On the Non-Specificity of Statistical Measures Used in Script Decipherment

arXiv:2608.02999v1 Announce Type: new Abstract: Statistical regularities are routinely offered as evidence that undeciphered sign systems encode language; the Indus script debate is the canonical exam

OPTD: On-Policy Transition Distillation with Consistency-Guided Adaptive Compression for Few-Step Diffusion Language Models

arXiv:2608.02942v1 Announce Type: new Abstract: Diffusion language models (dLLMs) can predict many tokens in parallel, but accurate generation still requires many iterative denoising steps. Few-step d

PAMT: Process-Aligned Reinforcement Learning for Multi-Domain Machine Translation

arXiv:2608.03077v1 Announce Type: new Abstract: Multi-domain machine translation (MDMT) requires more than fluent generation: it demands domain-sensitive translation decisions such as domain disambigu

ParVL: Parallel Scaling and Expandable Compute Allocation for Multimodal LLMs

arXiv:2608.04010v1 Announce Type: cross Abstract: Existing scaling strategies for Multimodal Large Language Models (MLLMs) typically expand either model parameters or sequential inference computation,

PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

arXiv:2608.04003v1 Announce Type: new Abstract: Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents offer a concrete setting for s

PDD-RRG: Posterior Diagnostic Decision for Study-level Radiology Report Generation

arXiv:2608.03055v1 Announce Type: cross Abstract: Automatic radiology report generation (RRG) aims to simulate the workflow of radiologists, assisting them in clinical diagnosis. However, existing met

Pingala: Prosody-Aware Decoding for Sanskrit Poetry Generation

arXiv:2603.24413v2 Announce Type: replace Abstract: Poetry generation in Sanskrit typically requires the verse to be semantically coherent and adhere to strict prosodic rules. In Sanskrit prosody, eve

Predicting Deep Neural Network Training Outcomes from Early Training Telemetry

arXiv:2608.03709v1 Announce Type: new Abstract: Large hyperparameter sweeps for deep neural networks spend substantial compute on configurations that are effectively doomed from the first few epochs.

Predicting Multilingual Classification and Translation Performance of LLMs with Cross-Lingual Alignment nicode{x2013} Is English Enough?

arXiv:2608.03446v1 Announce Type: new Abstract: Multilingual large language models (LLMs) have been shown to perform better on non-English classification tasks when the representations of the given la

Local Ai

Probing Character-level Transformers for the Spanish L-shaped Morphome

arXiv:2608.03452v1 Announce Type: new Abstract: When a transformer learns an irregular morphological pattern, what has it learned? Our test case is the Spanish L-shaped morphome, a complex irregular p

local-aiarxiv-cs-cl

Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning

arXiv:2608.02831v1 Announce Type: cross Abstract: Audio reasoning is essential for machine understanding of the acoustic world. Reinforcement learning with verifiable rewards can elicit such reasoning

Relational Priors as Convergence Pressure in LLM-Based Multi-Agent Systems

arXiv:2608.03239v1 Announce Type: new Abstract: Large language model-based multi-agent systems (LLM-MAS) are designed through roles, debate protocols, and aggregation rules. These choices create impli

Scalable Frequency- and Length-Aware Subdocument Deduplication for Large Language Model Pretraining

arXiv:2608.03089v1 Announce Type: new Abstract: Large-scale pretraining corpora contain substantial duplicate content. Although document-level deduplication is widely used, removing subdocument-level

Sensitivity, Causality, and Repair Dissociate: A Layer-Wise Analysis of Perturbation Robustness and Its Scaling

arXiv:2608.03842v1 Announce Type: new Abstract: When a language model fails on surface-perturbed input (typos, OCR noise, homophones), 'which layer is responsible' has three natural operationalization

SeqLLM: Augmenting LLMs with Behavioral-Sequence Modeling for High-Stakes Decisions at WeChat Pay

arXiv:2608.03063v1 Announce Type: new Abstract: Merchant risk control at large payment platforms screens tens of millions of merchants daily, where false positives harm legitimate merchants and false

SFT Conflicts, RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLMs

arXiv:2608.03573v1 Announce Type: new Abstract: Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) exhibit fundamentally different behaviors in enhancing multi-task reasoning for large langu

SimulRAG: Simulator-based RAG for Grounding LLMs in Long-form Scientific QA

arXiv:2509.25459v2 Announce Type: replace Abstract: Large Language Models (LLMs) show promise in generating long-form scientific explanations that synthesize evidence and connect multiple factors. How

SocietyBench: Forecasting Counterfactual Social-World Evolution

arXiv:2608.04009v1 Announce Type: new Abstract: Large language models (LLMs), and the agents built on top of them, are now benchmarked heavily on whether they can finish a task -- fix a bug, drive a b

Tutorials

Sparse Weight Decomposition for Efficient Circuit Extraction

arXiv:2608.03913v1 Announce Type: cross Abstract: Dense pretrained transformers do not naturally expose interpretable units for circuit extraction. Existing approaches obtain such units by learning au

tutorialsarxiv-cs-cl

States Hidden in Hidden States: Implicit Discrete State Representations Emerge in LLMs' Hidden States

arXiv:2407.11421v2 Announce Type: replace Abstract: Large Language Models (LLMs) exhibit emergent abilities that may reveal aspects of their internal mechanisms. We study one such capability: directly

string2string Studio: An Interactive, In-Browser Platform for String-to-String Algorithms

arXiv:2608.03984v1 Announce Type: new Abstract: We present string2string Studio, an interactive in-browser platform for string-to-string analysis across natural language processing, computational biol

Stuck on 'A': Diagnosing and Repairing Interface Injury in Attention-to-KDA Linearization of a 0.6B Language Model

arXiv:2608.02689v1 Announce Type: new Abstract: We convert 21 of 28 full-attention layers of Qwen3-0.6B-Base into KDA (Kimi Delta Attention) linear-attention layers on a single consumer-grade GPU budg

Suffix-Constrained Greedy Search Algorithms for Causal Language Models

arXiv:2603.01243v2 Announce Type: replace Abstract: Large language models (LLMs) are powerful tools that have found applications beyond human-machine interfaces and chatbots. Beside free-form generati

TabletCraft: Bridging a 4,000-Year Cultural Gap with Bidirectional Akkadian NMT and Cuneiform Rendering

arXiv:2608.02609v1 Announce Type: new Abstract: Half a million cuneiform clay tablets survive in museums worldwide, yet modern users can neither read nor write in the world's oldest writing system, le

Temporal Leakage in LLM Backtesting: Measurement, Validation, and Adjusted Scores

arXiv:2608.02985v1 Announce Type: cross Abstract: The standard check for contamination in LLM backtests is simple: compare scores before and after the training cutoff. We show this check is uninformat

Applications

The Eloquence team submission for task 1 of MLC-SLM challenge

arXiv:2507.19308v2 Announce Type: replace-cross Abstract: In this paper, we present our studies and experiments carried out for the task 1 of the Challenge and Workshop on Multilingual Conversational

applicationsarxiv-cs-cl

VetScore: Risk-Weighted Fact Verification for Veterinary Long-Form QA with Citations

arXiv:2608.03675v1 Announce Type: new Abstract: Citation excerpts can be used to increase the reliability of generated outputs and their faithfulness to cited sources, which is especially important in

VIVID: A Culturally Grounded Benchmark Exposing the Figurative Language Gap in Vietnamese NLP

arXiv:2608.03095v1 Announce Type: new Abstract: We present VIVID (Vietnamese Idioms for Validation and Interpretation Depth), the first systematic benchmark for evaluating culturally grounded figurati

What Language Does and What the Evidence Supports: A Functional Role Taxonomy and Evidence Audit of Language Grounding in Embodied Agents

arXiv:2608.03099v1 Announce Type: new Abstract: Foundation models place language throughout embodied agents, but its presence does not show what it contributes or how well that contribution is grounde

When Agents Learn to Be You: Benchmarking Privacy Leakage, Impersonation Risk, and Defenses in Persona Skills

arXiv:2608.03700v1 Announce Type: cross Abstract: Persona skills distill personal interaction histories into portable and executable artifacts for downstream agents. While enabling flexible personaliz

When Attention Goes Blind: Numerical Failure in ALiBi Positional Encodings

arXiv:2608.03994v1 Announce Type: new Abstract: We identify a previously overlooked failure mode of ALiBi positional encoding: its linear bias scaling underflows floating-point precision, which zeroes

Word Recovery in Large Language Models Enables Character-Level Tokenization Robustness

arXiv:2603.10771v2 Announce Type: replace Abstract: Large language models (LLMs) trained with canonical tokenization exhibit surprising robustness to non-canonical inputs such as character-level token

WorldCup Arena: Prospective, Leakage-Free Evaluation of Frontier LLMs on a Live Tournament

arXiv:2608.04008v1 Announce Type: new Abstract: Benchmarks that measure the forecasting ability of large language models are almost always retrospective: the event has happened, the answer is somewher

A Comprehensive FP8 Training Recipe for Reasoning-Enhanced Language Models

arXiv:2509.22536v5 Announce Type: replace Abstract: The immense computational cost of training Large Language Models (LLMs) presents a major barrier to innovation. While FP8 training offers a promisin

A Constitution-Grid Instrument for Data-Efficient RL Alignment (C-Guard)

arXiv:2608.00180v1 Announce Type: new Abstract: Conflicting objectives are general in RL alignment, and training on them data-efficiently is hard. Training a safety guard with RL means optimizing two

Agents

A False Average: Chain-of-Thought Monitors Collapse Where They Are the Only Defense

arXiv:2608.00583v1 Announce Type: cross Abstract: Chain-of-thought (CoT) monitoring is meant to catch the reward hacks that look clean in the actions and betray themselves only in the reasoning. We sh

agentsarxiv-cs-cl

A Few Neurons Reveal When LLMs Misuse Tools: Sparse Detection and Selective Steering for Reliable Tool Use

arXiv:2608.00218v1 Announce Type: new Abstract: Agentic LLMs exhibit three consequential tool-use failures: invalid arguments (validity), unnecessary calls (over-calling), and omitted calls when tools

Hardware

A Fortran General-Purpose Transpiler: Proof of Concept

arXiv:2608.00130v1 Announce Type: cross Abstract: Fortran has been the cornerstone of high-performance computing for decades and remains unmatched in many domains. Yet the language faces an expertise

hardwarearxiv-cs-cl

A Heuristic Perspective on Debiasing Language Models

arXiv:2608.00622v1 Announce Type: new Abstract: Language models (LMs) often acquire various biases during pre-training and may express them in interactions, potentially causing social harm. Existing m

A Large-Scale Multi-Dimensional Empirical Study of LLMs for Conversation Summarization

arXiv:2606.15974v2 Announce Type: replace Abstract: Despite the significant advancement of LLMs in conversation summarization, their evaluation remains limited by insufficient scenarios, input lengths

A Triple-Robustness Analysis of Retrieval-Augmented Generation for Multi-Hop Requirements Traceability

arXiv:2608.00705v1 Announce Type: cross Abstract: Reported verdicts on GraphRAG versus vector RAG disagree, and the evidence is typically tied to a single corpus, embedder, and judge -- and, we show,