·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
AutoSynthData: Generating Training Data for Enterprise Agents2h◆On the (In)effectiveness of AMR Augmentation for Large Language Models2h◆cua-speedrun: Standardized Benchmarking of the Speed of Computer-Use Agents2h◆Mitigating Memorization In Language Models2h◆MoEless: Efficient MoE LLM Serving with Serverless Experts2h◆Fork-Think with Confidence2h◆A Moving-Horizon Approximate Branch-and-Reduce Method for Deep Classification Trees2h◆Bongard: Training Machine Intuition2h◆Inference Auctions2h◆Semifactual Credit-Augmented Policy Optimization2h◆DEdit: Iterative Draft Editing for Speculative Decoding2h◆4MT-VLM: How Coarse Is a VLMs Cognitive Map?2h◆JuryFlow: Disagreement-Guided Human-in-the-Loop Multi-Agent Evaluation2h◆OverdoseMoE: A Multi-Expert Framework for Opioid Overdose Risk Prediction2h◆Values as Style: Disentangling Values from Semantics with One-Way Mixing for Low-Damage LLM Steering2h◆OverForge: Reasoning Through Strategies and Tactics Helps Cooperative Lifelong Adaptation2h◆Structure of Basic Human Values in Russian Social Media2h◆Superficial Reflection or Genuine Thought? A Fine-Grained Cognitive Analysis of Large Reasoning Models2h◆Evaluating Persistent Calibration under Evolving Model Knowledge2h◆Unmerge: Efficient Machine Unlearning via Task Arithmetic2h◆AutoSynthData: Generating Training Data for Enterprise Agents2h◆On the (In)effectiveness of AMR Augmentation for Large Language Models2h◆cua-speedrun: Standardized Benchmarking of the Speed of Computer-Use Agents2h◆Mitigating Memorization In Language Models2h◆MoEless: Efficient MoE LLM Serving with Serverless Experts2h◆Fork-Think with Confidence2h◆A Moving-Horizon Approximate Branch-and-Reduce Method for Deep Classification Trees2h◆Bongard: Training Machine Intuition2h◆Inference Auctions2h◆Semifactual Credit-Augmented Policy Optimization2h◆DEdit: Iterative Draft Editing for Speculative Decoding2h◆4MT-VLM: How Coarse Is a VLMs Cognitive Map?2h◆JuryFlow: Disagreement-Guided Human-in-the-Loop Multi-Agent Evaluation2h◆OverdoseMoE: A Multi-Expert Framework for Opioid Overdose Risk Prediction2h◆Values as Style: Disentangling Values from Semantics with One-Way Mixing for Low-Damage LLM Steering2h◆OverForge: Reasoning Through Strategies and Tactics Helps Cooperative Lifelong Adaptation2h◆Structure of Basic Human Values in Russian Social Media2h◆Superficial Reflection or Genuine Thought? A Fine-Grained Cognitive Analysis of Large Reasoning Models2h◆Evaluating Persistent Calibration under Evolving Model Knowledge2h◆Unmerge: Efficient Machine Unlearning via Task Arithmetic2h◆
News/Semifactual Credit-Augmented Policy Optimization
arxiv
PublishedOctober 2, 2026 at 4:00 AM
—neutral

Semifactual Credit-Augmented Policy Optimization

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

arXiv:2609.40360v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrelevant prompt features. We investigate this sensitivity through semifact

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
arxivOn the (In)effectiveness of AMR Augmentation for Large Language Models2harxivcua-speedrun: Standardized Benchmarking of the Speed of Computer-Use Agents2harxivMitigating Memorization In Language Models2harxivMoEless: Efficient MoE LLM Serving with Serverless Experts2h
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 ↗
Built by Marouane Gazouzi
HomeModelsNews