© 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. 128169
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on uppercase identifier-like tokens (e.g. acronyms, constants, and macros) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    Service Affinity, Target Architecture

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Descriere
    -1.06
    適量
    -1.04
    ek
    -0.92
    anz
    -0.91
     benzer
    -0.91
    ys
    -0.90
    ron
    -0.88
    뭘
    -0.87
    ion
    -0.86
    mel
    -0.85
    POSITIVE LOGITS
    semen
    1.15
     tertentu
    1.12
     ngờ
    0.96
    leopard
    0.95
     Risiko
    0.95
     bőr
    0.93
     tört
    0.93
     pág
    0.93
    HttpPut
    0.91
    octopus
    0.91
    Activations Density 0.010%

    No Known Activations