·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
SFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning4h◆Optical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning4h◆Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models4h◆Temporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents4h◆Why Muon Outperforms Adam: A Curvature Perspective4h◆Provably Auditable and Safe LLM Agents from Human-Authored Ontologies4h◆Efficient Reasoning on the Edge4h◆What Type of Inference is Active Inference?4h◆AutoLab: Can Frontier Models Solve Long-Horizon Auto Research and Engineering Tasks?4h◆MIRAGE: Mobile Agents with Implicit Reasoning and Generative World Models4h◆Inference-Time Vulnerability Beyond Shallow Safety: Alignment Along Generation Trajectories4h◆Characterizing initial human-AI proof formalization workflows4h◆AICompanionBench: Benchmarking LLMs-as-Judges for AI Companion Safety4h◆Strabo: Declarative Specification and Implementation of Agentic Interaction Protocols4h◆Knowledge Index of Noah's Ark4h◆AI from concrete to abstract: demystifying artificial intelligence to the general public4h◆How do machines learn? Evaluating the AIcon2abs method4h◆The Invisible Lottery: How Subtle Cues Steer Algorithm Choice in LLM Code Generation4h◆POLARIS: Guiding Small Models to Write Long Stories4h◆The Differentiable Auditory Loop (DAL): An ML Framework for Hyper-Personalized Hearing Aids4h◆SFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning4h◆Optical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning4h◆Dynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models4h◆Temporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents4h◆Why Muon Outperforms Adam: A Curvature Perspective4h◆Provably Auditable and Safe LLM Agents from Human-Authored Ontologies4h◆Efficient Reasoning on the Edge4h◆What Type of Inference is Active Inference?4h◆AutoLab: Can Frontier Models Solve Long-Horizon Auto Research and Engineering Tasks?4h◆MIRAGE: Mobile Agents with Implicit Reasoning and Generative World Models4h◆Inference-Time Vulnerability Beyond Shallow Safety: Alignment Along Generation Trajectories4h◆Characterizing initial human-AI proof formalization workflows4h◆AICompanionBench: Benchmarking LLMs-as-Judges for AI Companion Safety4h◆Strabo: Declarative Specification and Implementation of Agentic Interaction Protocols4h◆Knowledge Index of Noah's Ark4h◆AI from concrete to abstract: demystifying artificial intelligence to the general public4h◆How do machines learn? Evaluating the AIcon2abs method4h◆The Invisible Lottery: How Subtle Cues Steer Algorithm Choice in LLM Code Generation4h◆POLARIS: Guiding Small Models to Write Long Stories4h◆The Differentiable Auditory Loop (DAL): An ML Framework for Hyper-Personalized Hearing Aids4h◆
News/Beyond Log Likelihood: Probability-Based Objectives for Supervised Fine-Tuning across the Model Capability Continuum
arxiv
PublishedMay 25, 2026 at 4:00 AM
—neutral

Beyond Log Likelihood: Probability-Based Objectives for Supervised Fine-Tuning across the Model Capability Continuum

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

arXiv:2510.00526v3 Announce Type: replace Abstract: Supervised fine-tuning (SFT) is the standard approach for post-training large language models (LLMs), yet it often shows limited generalization. We trace this limitation to its default training objective: negative log likelihood (NLL). While NLL is

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
arxivSFMambaNet: Spectral-Frequency Enhanced Selective State Space Model for Correspondence Pruning4harxivOptical-Guided Neural Collapse for SAR Few-Shot Class Incremental Learning4harxivDynamic Infilling Anchors for Format-Constrained Generation in Diffusion Large Language Models4harxivTemporal Order Matters for Agentic Memory: Segment Trees for Long-Horizon Agents4h
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