© 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. 33299
    Prev
    Next
    INDEX
    Explanations

    zip files and codes

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

    The neuron fires whenever it sees the sequence “zip,” whether in file-compression contexts (zip/unzip, .zip files) or postal “Zip” code mentions.

    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
    ⁛
    -2.89
    躥
    -2.69
    ↵↵
    -2.56
    ⮑
    -2.50
     accommodating
    -2.44
    曏
    -2.44
     distinguishing
    -2.41
    
    -2.39
     extremely
    -2.39
    祤
    -2.38
    POSITIVE LOGITS
     was
    3.20
     Если
    2.84
     démocr
    2.39
    зывают
    2.30
    geſ
    2.27
    just
    2.25
    gebnisse
    2.23
    佧
    2.19
     были
    2.16
     عباس
    2.13
    Activations Density 0.006%

    No Known Activations