© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 112925
    Prev
    Next
    INDEX
    Explanations

    code file suffixes

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

    fragments of proper nouns or brand names in code, particularly those containing "Lite", "ache", "ico", or similar syllable patterns.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron detects occurrences of library or framework names (i.e. proper nouns from import/using statements) 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_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
     but
    -1.39
    (
    -1.27
     so
    -1.27
    ..."
    -1.23
     before
    -1.23
    ...
    -1.21
     I
    -1.20
     or
    -1.17
     and
    -1.16
    .....
    -1.14
    POSITIVE LOGITS
    違います
    1.40
    已经被
    1.34
    ление
    1.33
    antaine
    1.27
    öffnet
    1.27
    acje
    1.21
    ことです
    1.20
     Premios
    1.19
     discípulos
    1.19
     ネイビー
    1.19
    Activations Density 0.016%

    No Known Activations