arXiv Machine Learning
Aug 11
Depth-adaptive Inference of Looped Language Models via Continuous Depth Batching
arXiv:2608. 09444v1 Announce Type: new Abstract: A main promise of looped language models (LMs) is depth-adaptive inference.
Large language models
Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv AI
Aug 11
BDH-CQ: In-Context Learning with Recurrent Latent Reasoning
arXiv:2608. 09888v1 Announce Type: cross Abstract: We introduce BDH-CQ, a reasoning model that combines in-context learning with recurrent latent reasoning.
arXiv AI
Aug 11
Emergence Invariance: From Symbolized Thought to Structural Control
arXiv:2608. 01548v2 Announce Type: replace Abstract: Language-first intelligence is constrained by which distinctions enter its symbolic record, which mappings its language--interpreter--environment complex can execute, and which possibilities can be realized with finite resources.
arXiv AI
Aug 11
Enhancing AI Interpretability with Localised Architectures
arXiv:2606. 07998v3 Announce Type: replace-cross Abstract: Recent advances in generative AI, especially powerful Large Language Models (LLMs), raise concerns over the interpretability, safety and sustainability of these large and opaque AI models.
arXiv Machine Learning
Aug 11
Beyond the Capability Boundary: Zeroth-Order Optimization for Self-Evolving LLM Agents
arXiv:2608. 09292v1 Announce Type: new Abstract: Self-evolving methods improve the capabilities of LLM agents by sampling trajectories from the underlying LLMs and learning from these trajectories.
arXiv AI
Aug 11
Can We Optimize the Performance-Carbon Emission Break-Even Point?: The Quest for Greener LLMs
arXiv:2608. 08744v1 Announce Type: cross Abstract: The carbon footprint of any deployed Large Language Model (LLM) accumulates during inference, where repeated use of the model substantially exceeds the one-time cost of fine-tuning.
arXiv AI
Aug 11
STAIR: Effective Incident Response Using an End-to-End Agentic Planning Framework
arXiv:2608. 09524v1 Announce Type: cross Abstract: Incident response planning is critical for restoring compromised software systems after cyberattacks.
arXiv AI
Aug 11
Rethinking Factor Sharing in Federated LoRA: A Rank-Aware Adaptive Approach
arXiv:2608. 09742v1 Announce Type: cross Abstract: Low-rank adaptation (LoRA) represents large language model (LLM) updates with two compact matrix factors, i.
arXiv AI
Aug 11
MonitorBench: A Comprehensive Benchmark for Chain-of-Thought Monitorability in Large Language Models
arXiv:2603. 28590v3 Announce Type: replace Abstract: Large language models (LLMs) can generate chains of thought (CoTs) that are not always causally responsible for their final outputs.
arXiv AI
Aug 11
See Me, Believe Me: Causality, Intersectionality, and Interventions Improving the Appearance of Patients
arXiv:2410. 01227v2 Announce Type: replace-cross Abstract: In the context of medical records, patients often experience testimonial injustice, where the textual account undermines the validity of their experiences.
arXiv AI
Aug 11
Contamination Means Overestimation? A Fine-Grained Empirical Study in Code Intelligence
arXiv:2506. 02791v4 Announce Type: replace-cross Abstract: In recent years, code intelligence has gained increasing importance in the field of automated software engineering.
arXiv Machine Learning
Aug 11
Analogical Learning for Cross-Scenario Generalization: Framework and Application to Intelligent Localization
arXiv:2504. 08811v3 Announce Type: replace Abstract: Modern learning systems often struggle with joint learning across diverse scenarios and immediate adaptation to new ones, because they rely heavily on the scenario-dependent absolute data-label representations.
arXiv AI
Aug 11
Attack Ensembles Expose a Safety-Utility Trade-off in Black-Box Guard Defenses Against Encoded VLM Jailbreaks
arXiv:2607. 26574v2 Announce Type: replace-cross Abstract: Safety classifiers ("guards") are the dominant black-box defense for vision-language models, yet a guard judges an input's surface form, not its meaning: a harmful request re-encoded as set theory, formal logic, a classical language, code, or text rendered inside an image slips past a guard that would block it in plain language - the decode gap.
arXiv Machine Learning
Aug 11
ORBIT: Training-Free Multi-Attribute Behavioral Steering via Orthogonal Subspace Rotation
arXiv:2606. 22357v2 Announce Type: replace-cross Abstract: Language models are widely used in assistant settings, where controlling behavioral attributes is often essential.
arXiv AI
Aug 11
Proxy OPD: On-Policy Distillation with Transferable Relative Proxy Update
arXiv:2607. 11505v2 Announce Type: replace-cross Abstract: Post-training for large language models typically couples policy exploration with model optimization, hindering the reuse of high-reward behaviors from policy exploration.
arXiv AI
Aug 11
In-Situ Behavioral Evaluation for LLM Fairness, Not Standardized-Test Scores
arXiv:2605. 12530v2 Announce Type: replace-cross Abstract: LLM fairness should be evaluated through in-situ behavioral pattern rather than standardized-test Q&A benchmarks.
arXiv Machine Learning
Aug 11
Math-Vision Diagrams: A Comprehensive Benchmark for Evaluating LLM Mathematical Diagram Generation Capabilities
arXiv:2608. 08964v1 Announce Type: new Abstract: The generation of mathematically precise diagrams from tex- tual prompts has emerged as a critical yet underexplored capability of Large Language Models (LLMs).
arXiv Machine Learning
Aug 11
In-Context Density Estimation for Tabular Data
arXiv:2608. 09348v1 Announce Type: new Abstract: Density estimation underlies many unsupervised tasks on tabular data such as anomaly detection, out-of-distribution detection, and data augmentation.
arXiv AI
Aug 11
SkillsMetric: Mapping the Detection Boundary of Static Analysis for Malicious Agent Skills
arXiv:2608. 08468v1 Announce Type: cross Abstract: Agent Skills---structured packages of instructions and scripts that augment LLM-based agents---are rapidly proliferating, yet their security properties remain under-explored.
arXiv AI
Aug 11
LegoLM: Structured Weight Sharing for Large Language Models
arXiv:2608. 08652v1 Announce Type: cross Abstract: We present \LegoLM{}, a structured weight-sharing compression framework for large language models grounded in a systematic study of why global weight sharing fails and how to fix it.