mlsanityData to 13 Oct 2026

Different Feedback, Different Updates: Selective Self-Learning from User Interactions for Large Language Models

arXiv 2608.09109 ↗First posted 10 Aug 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

User feedback offers natural supervision for persistent LLM improvement, but a single message may support multiple behavioral changes with different scopes of generalization. We introduce SLIFT, a selective self-learning framework built on a task-relative view of user feedback. SLIFT decomposes each feedback message into atomic components and interprets each component relative to the original task as Fix, Spec, or Null: requirements for task validity, compatible condition-specific refinements, or content with no reliable positive update direction. To incorporate each change at the appropriate scope, SLIFT trains two complementary LoRA adapters on a shared frozen backbone: a Generalist that consolidates Fix requirements into default behavior through feedback-conditioned self-distillation, and a Specialist that observes only the task and Generalist response to supply residual guidance for applicable, unmet Spec refinements. Null components induce no positive update. Across backbones, SLIFT achieves strong performance on both MemoryBench and WildFB, with targeted analyses further examining its underlying mechanisms. We release our code at https://anonymous.4open.science/r/SLIFT.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Builds on 21 of its 39 references that are in the map. First posted on arXiv 10 Aug 2026.

Builds on

21 of 39 references.

References · 21

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,231
  2. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,799
  3. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,222
  4. 2025A-MEM: Agentic Memory for LLM Agentscited by 1,032
  5. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 877
  6. 2024WildChat: 1M ChatGPT Interaction Logs in the Wildcited by 616
  7. 2023Chain of Hindsight Aligns Language Models with Feedbackcited by 147
  8. 2025MemOS: A Memory OS for AI Systemcited by 134
  9. 2025Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memorycited by 134
  10. 2023RL4F: Generating Natural Language Feedback with Reinforcement Learning for Repairing Model Outputscited by 105
  11. 2026Ministral 3cited by 103
  12. 2025MemoryBench: A Benchmark for Memory and Continual Learning in LLM Systemscited by 59
9 more
  1. 2024WildFeedback: Aligning LLMs With In-situ User Interactions And Feedbackcited by 31
  2. 2022Learning to Revise References for Faithful Summarizationcited by 28
  3. 2025User Feedback in Human-LLM Dialogues: A Lens to Understand Users But Noisy as a Learning Signalcited by 19
  4. 2025Language Models Can Learn from Verbal Feedback Without Scalar Rewardscited by 17
  5. 2024Retrospective Learning from Interactionscited by 6
  6. 2026Beyond Experience Retrieval: Learning to Generate Utility-Optimized Structured Experience for Frozen LLMscited by 5
  7. 2025HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworkscited by 3
  8. 2026MulFeRL: Enhancing Reinforcement Learning with Verbal Feedback in a Multi-turn Loopcited by 3
  9. 2026WildReward: Learning Reward Models from In-the-Wild Human Interactionscited by 3

Built on by

Not cited yet.