© 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. 34-GEMMASCOPE-RES-131K
    7. 2800
    Prev
    Next
    INDEX
    Explanations

    API service context and events

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

    The neuron is selectively activating on programming‐style tokens—API and class names, import statements, code punctuation, and other typical code 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_34/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
    hkan
    -0.81
    したいと思います
    -0.81
    oplas
    -0.80
    FormBorderStyle
    -0.80
    िंग
    -0.77
     aortic
    -0.76
    liage
    -0.75
     elsewhere
    -0.75
     realising
    -0.74
    rika
    -0.73
    POSITIVE LOGITS
     work
    0.91
     worker
    0.90
     Constraints
    0.89
    Work
    0.87
    Constraints
    0.87
     Constraint
    0.85
     constraints
    0.84
     Work
    0.83
     workers
    0.80
     Periodic
    0.79
    Activations Density 0.006%

    No Known Activations