Unifying and Beyond the Theory of Pervasive Capability Degradation in LLM Post-training SFT from Feature Interference
cs.LG
Submitted: 2026-11-03
Updated: 2026-11-05
Code: https://github.com/Zodiark-ch/SFT-Conflict
Terminology
Sources
- Geometric Calibration of Supervised Fine-Tuning
- Finetuning with Sampling: SFT Learns Better Than You Think
- Data Mixing Optimization for Supervised Fine-Tuning of Large Language Models
- Laguerre Geometry for Interpreting Large Language Models
- LLM-Based Data Science Agents: A Survey of Capabilities, Challenges, and Future Directions
- Proximal Policy Optimization Algorithms
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Spurious Rewards Paradox: Mechanistically Understanding How RLVR Activates Memorization Shortcuts in LLMs
- Fine-Tuning Language Models from Human Preferences
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis