© 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. 40202
    Prev
    Next
    INDEX
    Explanations

    predisposing factors, incidental findings

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

    scientific terms in medical or research contexts.

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

    This neuron selectively activates on long, multi-syllable, morphologically complex “technical” words (i.e. those with prefixes, suffixes, or multiple morphemes).

    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
    appellant
    -1.39
    ượu
    -1.37
    黝
    -1.34
    cinoma
    -1.30
    者を
    -1.30
    ußer
    -1.30
     Ask
    -1.27
    ieson
    -1.27
     distanza
    -1.24
    cún
    -1.24
    POSITIVE LOGITS
    also
    1.82
    some
    1.49
    still
    1.42
     also
    1.41
     still
    1.40
    Also
    1.34
    $
    1.34
    橚
    1.30
    in
    1.30
    充電
    1.30
    Activations Density 0.060%

    No Known Activations