safety

Research Assistant: AstraZeneca's Agentic System for R&D

arXiv:2608.12395v1 Announce Type: new Abstract: We describe Research Assistant, an internal LLM-based system developed at AstraZeneca to help scientists and clinicians explore biomedical questions acr

Revisiting Overestimation Bias Problem of Q-learning: Settling Large Discrete Action Space via Action Intersection

Content type

All Blog X Post Paper YouTube Reddit GitHub

Clear filters

DGX agent

arXiv:2608.12912v1 Announce Type: new Abstract: This paper considers the overestimation bias problem of Q-learning in the setting of a large action space, for the purpose of relieving the bottleneck o

safetyarxiv-cs-lg

14 Aug 2026

Safe Exploration via Policy Priors

arXiv:2601.19612v4 Announce Type: replace-cross Abstract: Safe exploration is a key requirement for reinforcement learning (RL) agents to learn and adapt online, beyond controlled (e.g. simulated) env

Sampling Luck Masquerades as Allocation Gain: Auditing Test-Time Budget Allocation for Neural Combinatorial Optimization

arXiv:2608.13087v1 Announce Type: cross Abstract: Neural combinatorial optimization (NCO) solvers report the best of many sampled solutions per instance, and the sample count is, by convention, identi

Scaling Automatic Research Agents via World Models

arXiv:2608.12564v1 Announce Type: new Abstract: Automating empirical research is a long-standing direction of AI. Recent automatic research (AutoResearch) agents bring this goal within reach, as moder

SDAM: Structure-Difference-Aware Memory Evolution for Complex Text-to-SQL

arXiv:2608.12338v1 Announce Type: new Abstract: Text-to-SQL aims to convert natural language questions into executable SQL queries. While memory-based agent system improves complex SQL generation, exi

SDS-LoRA: Overcoming Anisotropic Gradient Scaling in Low-Rank Adaptation

arXiv:2606.16454v2 Announce Type: replace-cross Abstract: Low-Rank Adaptation (LoRA) enables efficient adaptation of large pretrained models to downstream tasks by parameterizing weight updates with l

SEAR: Sample Efficient Action Chunking Reinforcement Learning

arXiv:2603.01891v2 Announce Type: replace Abstract: Action chunking improves exploration and accelerates value propagation in long-horizon reinforcement learning, but naively applying off-policy metho

SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks

arXiv:2602.06854v2 Announce Type: replace Abstract: Multi-turn jailbreaks capture the real threat model for safety-aligned chatbots, where single-turn attacks are merely a special case. Yet existing a

Semantic Steering for Controllable Generation: Tuning-Free Concept Erasure in Multimodal Diffusion Transformers

arXiv:2608.12829v1 Announce Type: new Abstract: Multimodal Diffusion Transformers (MM-DiTs) have demonstrated remarkable text-to-image generation performance, surpassing traditional U-Net-based diffus

Simulation-to-real transfer learning for infrared spectroscopic chemical sensing and analysis from molecules to complex samples

arXiv:2608.13341v1 Announce Type: cross Abstract: Infrared (IR) spectroscopy is widely used for chemical sensing, but extracting reliable chemical information from spectra remains challenging. Convent

// Skill Misevolution in Self-Improving LLM Agents // Self-improving agents write their successes down as reusable skills. An unsafe success…

// Skill Misevolution in Self-Improving LLM Agents // Self-improving agents write their successes down as reusable skills. An unsafe success becomes reusable policy long after the input that triggered

safetydair-ai--x

SLAyiNG: A Diverse and Community-validated Dataset of Queer Slang

arXiv:2509.17449v2 Announce Type: replace Abstract: Queer vernacular is rarely studied in NLP, despite advancements in resources and evaluation for other sociolects and informal language. Because of t

SoK: From Generation to Consumption of Privacy Documents in Software Systems

arXiv:2608.12511v1 Announce Type: cross Abstract: Privacy documents (e.g., privacy policies) are a central mechanism through which digital services disclose data practices and seek user consent. Over

SpatialVAM:Spatial-Aware Multi-View Video Diffusion as a Data-Efficient Robot Policy

arXiv:2604.03181v2 Announce Type: replace-cross Abstract: Robotic manipulation requires understanding both the 3D spatial structure of the environment and its temporal evolution, yet most existing pol

StateBridge: Training-free Hidden-state Alignment for Latent Communication in LLM Multi-Agent Systems

arXiv:2608.13317v1 Announce Type: new Abstract: Large language model based multi-agent systems usually communicate in text, i.e., using discrete tokens. However, text introduces a discrete bottleneck.

Static analysis-guided agentic AI translation enables Rust as a full stack bioinformatics language

arXiv:2608.13029v1 Announce Type: cross Abstract: The field of bioinformatics struggles with legacy code - old code that is commonly used but may no longer have a maintainer, or may be written in an n

Sustaining Plasticity via Learnable Wavelet Activations in Continual Learning

arXiv:2608.12874v1 Announce Type: new Abstract: Plasticity loss has emerged as a critical challenge in continual learning that significantly hinders the acquisition of sequential tasks. While optimizi

Synthetic Persona Pretraining: Alignment from Token Zero

arXiv:2608.13482v1 Announce Type: cross Abstract: As language-model-based AI is increasingly deployed in autonomous settings, aligning its goals and values with those of humans becomes critical. Today

TANGCO: Learning Topology-Aware Capacity Allocation for Overload-driven Cascading Failures

arXiv:2608.13212v1 Announce Type: new Abstract: Networked systems, from power grids to traffic networks and cloud clusters, carry loads across nodes with limited capacity. A node whose load exceeds it

Teach the Magnitude, Not the Direction: Verifier-Bounded Credit Assignment for Multi-Turn Multi-step LLM Agents

arXiv:2608.13179v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) offers a verifier-bounded performance ceiling for training multi-turn tool-use agents, yet its tra

Temporal GRPO: Beyond Trajectory-Level Credit in Vision-Language-Action Reinforcement Learning

arXiv:2608.13026v1 Announce Type: new Abstract: Outcome-driven reinforcement learning offers a scalable way to post-train vision-language-action (VLA) policies from sparse task-success feedback. In co

safetyarxiv-cs-ro

that should be fairly obvious even now

that should be fairly obvious even now Looking back with a few year's perspective, it seems pretty likely to me that @hlntnr was correct that ending OpenAI due to firing CEO Sam Altman 'would actually

The Illusion of Improvement: Reject Inference Strategies in Credit Scoring

arXiv:2606.18479v2 Announce Type: replace Abstract: Reject inference methods are widely used to mitigate survival bias in credit scoring, yet their effectiveness remains poorly understood. We systemat

The Role of Natural Language Understanding in Multimodal Video-Based Dengue Diagnosis

arXiv:2608.12677v1 Announce Type: new Abstract: Detecting infection-related behavioral changes in mosquitoes from video data is challenging because mosquitoes are small, move rapidly and irregularly,

The Time Value of Evolution

arXiv:2608.13297v1 Announce Type: new Abstract: In evolutionary search, a weak child can be a valuable ancestor that makes high-fitness regions reachable. Immediate-return control is blind to this del

Thermodynamics of Learning: A Typed Four-Component Accounting of Memory, Fit, and Value

arXiv:2608.12791v1 Announce Type: cross Abstract: What a finite learning device has recorded and what will hold value for it on future tasks are not the same quantity. We develop a typed accounting fo

“This nightmare will only cost you eleventy trillion dollars…” Wild that to hear tech people speculate that the growing AI backlash is some …

“This nightmare will only cost you eleventy trillion dollars…” Wild that to hear tech people speculate that the growing AI backlash is some kind of “Chinese op” when, like, “we’re taking your job and

UniTexture: Cross-Task Universal Adversarial Textures for Vision-Language-Action Models

arXiv:2608.13453v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have emerged as generalist robotic policies capable of following diverse language instructions and performing a wi

Unmasking Conversational Bias in AI Multiagent Systems

arXiv:2501.14844v3 Announce Type: replace-cross Abstract: Detecting biases in the outputs produced by generative models is essential to reduce the potential risks associated with their application in

Unsupervised Deep Learning-based Keypoint Localization Estimating Descriptor Matching Performance

arXiv:2505.02779v2 Announce Type: replace Abstract: Retinal image registration, particularly for color fundus images, is a challenging yet essential task with diverse clinical applications. Existing r

Variance Reduction Based Experience Replay for Policy Optimization

arXiv:2602.05379v2 Announce Type: replace-cross Abstract: Effective reinforcement learning (RL) for complex stochastic systems requires leveraging historical data to improve sample efficiency and acce

vToken: Token-Level Virtualization for Reclaimable KV Caches

arXiv:2608.13263v1 Announce Type: new Abstract: Large language model serving faces a critical memory bottleneck: the KV cache grows with sequence length and batch size. PagedAttention uses fixed-size

When Should Multi-Round RAG Stop? Structured Stopping Judgments and Retrieval Reduction in Search-R1

arXiv:2608.13237v1 Announce Type: cross Abstract: Multi-round retrieval-augmented generation (RAG) must decide when to stop searching as evidence accumulates. Because the deployed policy is determined

Who Speaks Matters: Authority-Aware Multi-View RAG over Italian Parliamentary Proceedings

arXiv:2608.13410v1 Announce Type: new Abstract: Parliamentary proceedings are a primary record of democratic deliberation, yet their volume and fragmentation make multi-perspective access difficult fo

wow! new OpenAI study shows “no correlation between how often employees use AI and how much money they make. The ROI of the latest tech boom…

wow! new OpenAI study shows “no correlation between how often employees use AI and how much money they make. The ROI of the latest tech boom is still very much up in the air.” Let that sink in. This p

A-3PO: Accelerating Asynchronous LLM Training with Staleness-aware Proximal Policy Approximation

arXiv:2512.06547v4 Announce Type: replace-cross Abstract: Decoupled PPO has been a successful reinforcement learning (RL) algorithm to deal with the high data staleness under the asynchronous RL setti

A method for tissue-mask supported whole-body image registration in the UK Biobank

arXiv:2512.02702v3 Announce Type: replace Abstract: The UK Biobank is a large-scale study collecting imaging and non-imaging health data. Robust and accurate inter-subject image registration of these

A Modular Agentic Framework for Synthetically Constrained Multi-Objective Hit-to-Lead Optimization

arXiv:2608.11483v1 Announce Type: new Abstract: Hit-to-lead optimization requires iterative design of hit analogs across competing potency, selectivity, physicochemical, pharmacokinetic, safety, and s

A New First-Order Meta-Learning Algorithm with Convergence Guarantees

arXiv:2409.03682v2 Announce Type: replace Abstract: Learning new tasks by leveraging prior experience is a fundamental trait of intelligent systems. While Model-Agnostic Meta-Learning (MAML) is a lead

Accuracy and Order Sensitivity Diverge Under Label-Free Strategies

arXiv:2608.11947v1 Announce Type: cross Abstract: Multiple-choice benchmarks are widely used to evaluate large language models, but MCQ scores conflate knowledge with sensitivity to option order, whic

Adaptation of Generalist Robot Policies with Minimal Data

arXiv:2608.11363v1 Announce Type: cross Abstract: A central goal in robot learning is to move beyond task-specific human data collection toward robots that improve through autonomous interaction. Yet

AI Guardrail Survival under Single-Cycle Agentic Self-Summarization

arXiv:2608.11392v1 Announce Type: cross Abstract: Long-running agents periodically compact their context, replacing the transcript with a model-generated summary.Recent work shows that dropping a stan

Alignment of Similarity-Transformed Images Based on Fourier--Mellin Transform Using Auxiliary Function Method

arXiv:2608.11565v1 Announce Type: cross Abstract: This paper proposes an algorithm for estimating the similarity transformation, namely translation, scale, and rotation, between two images with subpix

anthropic’s best hope is that OpenAI falls apart. which it might.

anthropic’s best hope is that OpenAI falls apart. which it might. NEW from Ramp AI Index: disappointing adoption of Fable 5 We've heard several reasons from businesses...mainly Fable 5 is just too exp

Anti-Shortcut Distillation via Temporal Negative Knowledge Transfer

arXiv:2608.11789v1 Announce Type: new Abstract: Knowledge distillation (KD) trains a compact student by attracting it towards a converged teacher. It is silent about which directions the teacher itsel

AutoGrable: What Is a Good Graph for a Table?

arXiv:2608.11431v1 Announce Type: new Abstract: Graph learning presupposes a graph, and tables and relational databases do not come with one. Applying a GNN to them requires deciding which entities be

Benchmarking Trustworthiness of SLMs: Pre-trained vs. Compressed

arXiv:2608.11981v1 Announce Type: new Abstract: Small Language Models (SLMs) have emerged as a more efficient alternative to traditional Large Language Models (LLMs), offering promising potential in r