© 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. 33212
    Prev
    Next
    INDEX
    Explanations

    intramolecular

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

    square brackets in programming and query syntax.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron primarily triggers on numeric tokens (literal numbers) in code snippets.

    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
     because
    -2.08
     by
    -1.85
     from
    -1.64
     during
    -1.64
     even
    -1.57
     (
    -1.55
     as
    -1.48
     when
    -1.47
     but
    -1.44
     that
    -1.41
    POSITIVE LOGITS
     strikingly
    1.66
     amazingly
    1.38
     remarkably
    1.36
     diversos
    1.35
     horrid
    1.34
     marvelous
    1.34
     atteindre
    1.33
    ごとの
    1.33
     startling
    1.31
     joyful
    1.30
    Activations Density 0.002%

    No Known Activations