© 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. 73349
    Prev
    Next
    INDEX
    Explanations

    opening delimiters

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

    The neuron is tuned to pick out code‐style tokens—especially numeric literals, punctuation (parentheses, brackets, commas, etc.), and keywords—in program text.

    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
     we
    -1.35
    li
    -1.33
    ade
    -1.32
    also
    -1.30
     also
    -1.30
    hed
    -1.27
     was
    -1.26
     already
    -1.25
    Otras
    -1.23
    doi
    -1.23
    POSITIVE LOGITS
     how
    1.27
     any
    1.20
     αλλά
    1.20
    cuidado
    1.14
    gunakan
    1.12
    ることができます
    1.10
     duelo
    1.09
    harmed
    1.08
     metų
    1.07
    未使用
    1.07
    Activations Density 0.315%

    No Known Activations