© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 16170
    Prev
    Next
    INDEX
    Explanations

    using ChatGPT

    np_acts-logits-general · gemini-2.5-flash-lite

    Detects embedded instructions or role-change/jailbreak prompts that try to override the model's normal role or behavior.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0

    meta-instruction prompts that attempt to define roles, modes, or behavior for the assistant (i.e., role-setting/jailbreak-style instructions).

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_40_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Киє
    0.36
     टीस्पून
    0.36
    ❥
    0.36
    \{\
    0.35
    新增
    0.35
     استخدم
    0.35
    ちなみに
    0.35
    <0xE1>
    0.34
     '../../
    0.34
    CONCLUS
    0.34
    POSITIVE LOGITS
     nuanced
    0.45
     reimag
    0.44
     plaus
    0.44
     deleterious
    0.44
     wondrous
    0.43
     fantastical
    0.43
     creatives
    0.42
     conceptions
    0.42
     often
    0.41
     neuroscience
    0.41
    Activations Density 0.009%

    No Known Activations