© 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. 34-GEMMASCOPE-RES-131K
    4. 89780
    Prev
    Next
    INDEX
    Explanations

    PGP word list

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

    The neuron consistently activates on numeric tokens (digits and multi‐digit strings), including version numbers, counts, and similar numeral sequences.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Guidelines
    -0.77
    ostok
    -0.70
    granate
    -0.70
     Duckworth
    -0.69
    lishes
    -0.69
     at
    -0.68
    的には
    -0.68
    ਵ
    -0.68
     Rules
    -0.66
    Lambda
    -0.66
    POSITIVE LOGITS
     specif
    0.82
    דול
    0.81
     鸡
    0.81
     primiti
    0.79
     Lessing
    0.78
     pecho
    0.77
     vaccinations
    0.76
     conserving
    0.75
     corrig
    0.74
     fundament
    0.73
    Activations Density 0.005%

    No Known Activations