© 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. 99938
    Prev
    Next
    INDEX
    Explanations

    code identifiers and punctuation

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

    The neuron activates on capitalized or camel-cased tokens in code—names of classes, constructors, imported modules, or other type-level identifiers.

    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
    Setelah
    -1.53
    Setiap
    -1.38
    প্র
    -1.37
    霰
    -1.37
    Sebagai
    -1.36
    Kelebihan
    -1.35
    Menurut
    -1.34
    뜬
    -1.34
    なぜ
    -1.32
    nadequate
    -1.31
    POSITIVE LOGITS
     you
    2.22
     such
    2.02
     this
    1.57
     can
    1.48
     and
    1.40
     who
    1.38
    `.`
    1.37
    </sub>
    1.34
     suicidio
    1.30
    *',
    1.28
    Activations Density 0.048%

    No Known Activations