© 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. 36523
    Prev
    Next
    INDEX
    Explanations

    `i` followed by `new`, `require`, `fact`

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

    This neuron fires on programming‐style code tokens—marking segments of source code (e.g., function definitions, assignments, numeric literals) rather than natural-language 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_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
     epitom
    -1.16
     before
    -1.12
     even
    -1.10
     from
    -1.09
     if
    -1.07
     undoubtedly
    -1.02
     one
    -1.02
     give
    -1.02
     unrelenting
    -1.00
     brutal
    -1.00
    POSITIVE LOGITS
     BECAUSE
    1.06
     miesto
    1.05
     TWENTY
    1.04
     ì
    1.03
     ludicrous
    1.02
     decidedly
    1.02
     nonchal
    1.01
     grosser
    1.00
     majestic
    0.99
     amd
    0.98
    Activations Density 0.015%

    No Known Activations