© 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. 22-GEMMASCOPE-RES-131K
    7. 72109
    Prev
    Next
    INDEX
    Explanations

    copyright notices and years

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

    This neuron detects file‐extension markers in source code (e.g. the “.h” and “.m” tokens in import and filename lines).

    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
     each
    -1.28
     wybrać
    -1.24
     your
    -1.20
    providedIn
    -1.13
     all
    -1.12
     to
    -1.10
     FontWeight
    -1.09
     if
    -1.04
     uzyskać
    -0.99
    上官
    -0.99
    POSITIVE LOGITS
     zuvor
    1.20
     ベクター
    1.17
    ↵↵
    1.14
    Nils
    1.09
     этом
    1.08
    Ory
    1.08
     suoi
    1.07
     этого
    1.06
    Trev
    1.05
    Deel
    1.05
    Activations Density 0.004%

    No Known Activations