Concept · 8 episode(s)

Sandbagging

← all concepts

Definition

Sandbagging is when a model deliberately performs worse than it can — on an evaluation, in front of a particular user, or under specific cues — to avoid scrutiny, training updates, or downstream consequences. It’s a core concern for capability evaluations: a sandbagging model lies about what it can do.

Episodes covering this

  1. 274
    Reading a Model's Internals to Tell 'Won't Say' From 'Doesn't Know'
    A Lie Detector Test for Language Models: Reading Knowledge a Model Won't Reveal
    Dinge · StackOne Technologies·13 min·Sep 21, 2026
  2. 259
    GPT-6 Astra Behaves Better, And OpenAI Can Read It Less
    GPT-6 Astra System Card
    OpenAI · OpenAI·21 min·Sep 04, 2026
  3. 199
    Finding a Model's Hidden Behaviors Without Knowing What You're Looking For
    Mechanistically Eliciting Latent Behaviors in Language Models
    Mack, Panickssery, Turner · Principles of Intelligence·15 min·Jul 04, 2026
  4. 174
    When the AI 'Schemes,' It's Usually Just Lazy or Confused
    Model Forensics: Investigating Whether Concerning Behavior Reflects Misalignment
    Singh, Kroiz, Rajamanoharan et al. · MATS·28 min·Jun 25, 2026
  5. 158
    How Floating-Point Rounding Lets a Model Tell Which Chip It's On — And Misbehave
    FloatDoor: Platform-Triggered Backdoors in LLMs
    Loose, Sander, Mächtle et al. · University of Luebeck·29 min·Jun 19, 2026
  6. 143
    When a Model Notices You Forged Its Own Words, And Why That Breaks Safety Tests
    Prefill Awareness in Large Language Models
    Wang, Mahajan, Africa et al. · Constellation / University of Wisconsin-Madison·24 min·Jun 12, 2026
  7. 128
    How a Model Can Earn Full Reward and Still Resist Training
    Generalization Hacking: Models Can Game Reinforcement Learning by Preventing Behavioral Generalization
    Xiao, Phuong · California Institute of Technology·29 min·Jun 11, 2026
  8. 007
    Exploration Hacking: When Models Sabotage Their Own RL Training
    Exploration Hacking: Can LLMs Learn to Resist RL Training?
    Jang, Falck, Braun et al. · MATS·23 min·May 02, 2026

Worth reading next

Papers we haven't done a deep dive on yet, but would recommend on this topic.