© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14651
    Prev
    Next
    INDEX
    Explanations

    code prefixes and suffixes

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

    The neuron is picking out occurrences of library or framework identifiers (e.g. “ngx,” “Html,” “geom,” etc.) 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_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
     all
    -1.48
     in
    -1.47
     what
    -1.41
     even
    -1.32
    corbic
    -1.30
     which
    -1.27
    ｪ
    -1.25
    ]$.
    -1.20
    lijken
    -1.20
     one
    -1.16
    POSITIVE LOGITS
    Ꮣ
    1.19
     PERFECT
    1.19
     betreffende
    1.16
     canadian
    1.15
     kroppen
    1.13
     området
    1.12
    了许多
    1.09
    炱
    1.09
    værk
    1.09
     offensichtlich
    1.07
    Activations Density 0.031%

    No Known Activations