·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
GAGPO: Generalized Advantage Grouped Policy Optimization1h◆UP-NRPA: User Portrait based Nested Rollout Policy Adaptation for Planning with Large Language Models in Goal-oriented Dialogue Systems1h◆Orchestra-o1: Omnimodal Agent Orchestration1h◆Hybrid Open-Ended Tri-Evolution Makes Better Deep Researcher1h◆WorkBench Revisited: Workplace Agents Two Years On1h◆Refusal Beyond a Single Direction: A Preliminary Comparison of Diff-in-Means and INLP1h◆YeasierAgent: Agentic Social Sandbox as a Canvas for Intent-Driven Creation of Platform-Agnostic Symbiotic Agent-Native Applications1h◆MA-ProofBench: A Two-Tiered Evaluation of LLMs for Theorem Proving in Mathematical Analysis1h◆A Multi-Agent AI System for Automated High School Transcript Processing: Collaborative Document Analysis at Scale1h◆Closing the Reflection Gap: A Free Calibration Bonus for Agentic RL1h◆SkillAudit: Ground-Truth-Free Skill Evolution via Paired Trajectory Auditing1h◆CSPO: Constraint-Sensitive Policy Optimization for Safe Reinforcement Learning1h◆Towards Direct Latent-Space Synthesis for Parallel Branches in LLM-Agent Workflows1h◆An Agentic Retrieval Framework for Autonomous Context-Aware Data Quality Assessment1h◆Aligning Quantum Operators with Large Language Models1h◆AI can help scientists publish less1h◆Safety-Contract Graph Multi-Agent Reinforcement Learning for Autonomous Network Security Response1h◆When Plausible Is Not Realistic: Evaluating Human Mobility in LLM-Based Urban Simulation1h◆Rethinking Backdoor Adversarial Unlearning through the Lens of Catastrophic Forgetting in Continual Learning1h◆Clay-CNN Hybrids: Leveraging Geo-Foundational Models as Auxiliary Context for Landslide Detection1h◆GAGPO: Generalized Advantage Grouped Policy Optimization1h◆UP-NRPA: User Portrait based Nested Rollout Policy Adaptation for Planning with Large Language Models in Goal-oriented Dialogue Systems1h◆Orchestra-o1: Omnimodal Agent Orchestration1h◆Hybrid Open-Ended Tri-Evolution Makes Better Deep Researcher1h◆WorkBench Revisited: Workplace Agents Two Years On1h◆Refusal Beyond a Single Direction: A Preliminary Comparison of Diff-in-Means and INLP1h◆YeasierAgent: Agentic Social Sandbox as a Canvas for Intent-Driven Creation of Platform-Agnostic Symbiotic Agent-Native Applications1h◆MA-ProofBench: A Two-Tiered Evaluation of LLMs for Theorem Proving in Mathematical Analysis1h◆A Multi-Agent AI System for Automated High School Transcript Processing: Collaborative Document Analysis at Scale1h◆Closing the Reflection Gap: A Free Calibration Bonus for Agentic RL1h◆SkillAudit: Ground-Truth-Free Skill Evolution via Paired Trajectory Auditing1h◆CSPO: Constraint-Sensitive Policy Optimization for Safe Reinforcement Learning1h◆Towards Direct Latent-Space Synthesis for Parallel Branches in LLM-Agent Workflows1h◆An Agentic Retrieval Framework for Autonomous Context-Aware Data Quality Assessment1h◆Aligning Quantum Operators with Large Language Models1h◆AI can help scientists publish less1h◆Safety-Contract Graph Multi-Agent Reinforcement Learning for Autonomous Network Security Response1h◆When Plausible Is Not Realistic: Evaluating Human Mobility in LLM-Based Urban Simulation1h◆Rethinking Backdoor Adversarial Unlearning through the Lens of Catastrophic Forgetting in Continual Learning1h◆Clay-CNN Hybrids: Leveraging Geo-Foundational Models as Auxiliary Context for Landslide Detection1h◆
News/Refusal Beyond a Single Direction: A Preliminary Comparison of Diff-in-Means and INLP
arxiv
PublishedJune 15, 2026 at 4:00 AM
—neutral

Refusal Beyond a Single Direction: A Preliminary Comparison of Diff-in-Means and INLP

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2606.13720v1 Announce Type: new Abstract: Arditi et al. (2024) has shown that refusal in safety fine-tuned chat models is mediated by a single linear direction in the residual stream, recoverable by a difference-in-means (DiM) of harmful and harmless activations. We compare DiM-based intervent

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivGAGPO: Generalized Advantage Grouped Policy Optimization1harxivUP-NRPA: User Portrait based Nested Rollout Policy Adaptation for Planning with Large Language Models in Goal-oriented Dialogue Systems1harxivOrchestra-o1: Omnimodal Agent Orchestration1harxivHybrid Open-Ended Tri-Evolution Makes Better Deep Researcher1h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews