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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25520
    Prev
    Next
    INDEX
    Explanations

    statements of prohibition

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

    The neuron activates on prohibition or negation cues—words like “not,” “no,” “must,” or “may not” that signal forbidden actions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     lidé
    -1.24
     känner
    -1.23
    ayangkan
    -1.15
    遊んで
    -1.14
     faktiskt
    -1.11
     capuz
    -1.11
     biraz
    -1.09
    sofá
    -1.09
     márm
    -1.09
     ještě
    -1.09
    POSITIVE LOGITS
     use
    1.55
     any
    1.41
    any
    1.20
    Ÿ
    1.15
    ･･･
    1.15
    T
    1.14
     ever
    1.13
    使用
    1.12
     EVER
    1.08
    If
    1.05
    Activations Density 0.027%

    No Known Activations