© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 70672
    Prev
    Next
    INDEX
    Explanations

    after, before actions

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

    The neuron is highly sensitive to numeric and list‐style tokens—digits, Roman numerals, step numbers, and their accompanying punctuation or markers.

    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
    Deadpool
    -1.01
     torba
    -1.00
    sich
    -1.00
     Ел
    -0.99
     чого
    -0.98
     merawat
    -0.96
    bringt
    -0.96
     mendengarkan
    -0.96
     čier
    -0.95
    tlere
    -0.95
    POSITIVE LOGITS
     you
    1.15
    www
    1.15
     consider
    1.09
     www
    1.02
    Figure
    1.02
    E
    0.98
    Q
    0.96
     want
    0.95
    ri
    0.94
    od
    0.92
    Activations Density 0.008%

    No Known Activations