arXiv Machine Learning
Jun 30
Sparse Autoencoders are Capable LLM Jailbreak Mitigators
arXiv:2602. 12418v2 Announce Type: replace-cross Abstract: Jailbreak attacks remain a persistent threat to large language model safety.
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
Jun 30
IMCBench: A benchmark for multimodal LLMs in Image-grounded Medical Conversations
arXiv:2606. 28556v1 Announce Type: new Abstract: Recent advances in large language models and vision-language models have enabled reasoning over multimodal data, offering opportunities for clinical applications such as decision support and triaging.
arXiv Machine Learning
Jun 30
When Does Online Imitation Learning Help in LLM Post-Training? The Role of (Non-)Realizability Beyond Horizon
arXiv:2606. 30445v1 Announce Type: new Abstract: Online imitation learning (IL), particularly on-policy distillation, has emerged as a strong LLM post-training approach, often outperforming offline supervised fine-tuning (SFT).
arXiv AI
Jun 30
Diagnosing and Repairing Factual Errors in RAG under Budget Constraints
arXiv:2606. 29377v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) improves the factuality of large language models by grounding responses in external evidence, yet real-world deployments remain fragile.
arXiv AI
Jun 30
Towards Human-Level Book-Writing Capability
arXiv:2605. 17064v2 Announce Type: replace Abstract: Large language models are optimized for instruction following and agentic tasks remain poorly aligned with the requirements of high-quality creative writing.
arXiv Machine Learning
Jun 30
Depth Exploration for LLM Decoding
arXiv:2606. 29223v1 Announce Type: new Abstract: Autoregressive LLM decoding evaluates every generated token through the full layer stack, even though many tokens become predictable at intermediate depths.
arXiv AI
Jun 30
Mandol: An Agglomerative Agent Memory System for Long-Term Conversations
arXiv:2606. 29778v1 Announce Type: cross Abstract: Long-term conversational agents need to remember and query cross-session, multi-typed information with complex correlations.
arXiv AI
Jun 30
Steerable Visual Representations
arXiv:2604. 02327v2 Announce Type: replace-cross Abstract: Pretrained Vision Transformers (ViTs) such as DINOv2 and MAE provide generic image features that can be applied to a variety of downstream tasks such as retrieval, classification, and segmentation.
arXiv AI
Jun 30
Redefining Maritime Anomaly Detection via Equation-Grounded Synthetic Anomalies
arXiv:2606. 29721v1 Announce Type: cross Abstract: Maritime anomaly detection is essential for ensuring maritime safety, security, and efficient traffic management at sea, with Automatic Identification System (AIS) data serving as a primary data source.
arXiv AI
Jun 30
StarDojo: Benchmarking Open-Ended Behaviors of Agentic Multimodal LLMs in Production-Living Simulations with Stardew Valley
arXiv:2507. 07445v3 Announce Type: replace Abstract: Autonomous agents navigating human society must master both production activities and social interactions, yet existing benchmarks rarely evaluate these skills simultaneously.
arXiv AI
Jun 30
The Heterogeneous Safety Impacts of Benign Multilingual Fine-Tuning
arXiv:2606. 28843v1 Announce Type: cross Abstract: Fine-tuning a large language model is a ubiquitous method for enhancing its capability on a specific downstream task.
arXiv AI
Jun 30
Building to the Test: Coding Agents Deliver What You Check, Not What You Requested
arXiv:2606. 28430v1 Announce Type: cross Abstract: Benchmarks are widely used to evaluate task completion by Large Language Models (LLMs), but this approach has accumulated construction-validity problems, and a passing score may not show whether the requested task was delivered.
arXiv AI
Jun 30
Mechanistic Personality Analysis of LLMs Steering Personality via Latent Feature Interventions
arXiv:2606. 28770v1 Announce Type: new Abstract: Large Language Models (LLMs) have demonstrated the ability to simulate human-like OCEAN personality traits in generated text.
arXiv AI
Jun 30
SemJoin: Semantic Join Optimization
arXiv:2606. 29532v1 Announce Type: cross Abstract: Integrating unstructured data into relational database systems is increasingly important as demand grows for natural language querying and analysis.
arXiv AI
Jun 30
Online Data Selection for Instruction Tuning via Gaussian Processes
arXiv:2606. 30077v1 Announce Type: cross Abstract: With Large Language Model (LLM) pre-training and fine-tuning shifting its focus from data volume to data quality, quality data selection has emerged as a critical research topic.
arXiv AI
Jun 30
ImprovEvolve: Basin-Hopping Meets LLM-Guided Evolutionary Search
arXiv:2602. 10233v2 Announce Type: replace-cross Abstract: LLM-guided evolutionary computation, most notably AlphaEvolve, has been remarkably successful in discovering novel mathematical constructions by solving challenging optimization problems.
arXiv AI
Jun 30
Agent Safety Is Action Alignment
arXiv:2606. 28739v1 Announce Type: new Abstract: Large language models increasingly act as agents: they call tools, move money, delete records, and send messages on a user's behalf.
arXiv AI
Jun 30
Few-Shot Domain Incremental Learning via Continual Vision-Language Consolidation
arXiv:2606. 30190v1 Announce Type: cross Abstract: Existing domain-incremental learning (DIL) strategies call for massive amounts of data to adapt to new domains and suffer from the overfitting problem in the case of data scarcity.
arXiv AI
Jun 30
Spectral Perturbation of the Empirical Fisher Information Matrix under Weight Quantization
arXiv:2606. 28432v1 Announce Type: cross Abstract: We study the spectral perturbation of the empirical Fisher Information Matrix (FIM) of a parametric statistical model under two structured perturbations: departure of the input from a reference (in-distribution) ensemble, and finite-precision (quantized) perturbation of the model's parameters.
arXiv AI
Jun 30
SABER-Math: Automated Benchmark for Information Retrieval Evaluation in Mathematics
arXiv:2606. 29894v1 Announce Type: cross Abstract: As agentic AI systems tackle more complex mathematical tasks, they increasingly rely on information retrieval (IR) to search problem databases, theorem libraries, and educational resources.