© 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. 10-GEMMASCOPE-RES-131K
    7. 73080
    Prev
    Next
    INDEX
    Explanations

    crashing through barriers

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

    This neuron detects expressions of things slipping, falling, or sinking “through” gaps (e.g. crack(s), net, floor).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     and
    -1.43
    女優
    -1.36
     시간
    -1.20
    seminar
    -1.20
    flä
    -1.20
     her
    -1.20
     be
    -1.20
    窳
    -1.19
     Regiment
    -1.19
    少し
    -1.18
    POSITIVE LOGITS
    全是
    1.27
     into
    1.21
    .”
    1.20
    的态度
    1.13
    どく
    1.13
    之处
    1.09
    OP
    1.08
    べて
    1.08
     prüfen
    1.06
    .}
    1.05
    Activations Density 0.027%

    No Known Activations