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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 84486
    Prev
    Next
    INDEX
    Explanations

    introduces definitions, questions, or analyses

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

    The neuron fires on the token “Def,” i.e. when it sees a function‐ or method‐definition keyword in code.

    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
     TÉCN
    -1.45
     ensaio
    -1.40
     allerlei
    -1.38
     необходимость
    -1.33
    鼉
    -1.30
     tromper
    -1.30
     kompati
    -1.29
     taşlı
    -1.29
     klanten
    -1.27
     verhind
    -1.26
    POSITIVE LOGITS
     of
    2.30
     not
    1.59
     but
    1.46
     at
    1.43
     our
    1.35
     hábit
    1.30
    も
    1.30
     with
    1.26
     amidst
    1.26
     had
    1.22
    Activations Density 0.000%

    No Known Activations