© 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. 57265
    Prev
    Next
    INDEX
    Explanations

    code starts with ===

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

    The neuron strongly activates on unusually rare or unique tokens—such as proper‐name fragments, code delimiters (e.g. “===”), or other low-frequency strings.

    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
     opportunities
    -0.90
     Meisters
    -0.90
     церковь
    -0.74
    촌
    -0.73
     for
    -0.73
    äumt
    -0.72
    altra
    -0.71
     देखा
    -0.71
     contrast
    -0.70
    伝え
    -0.70
    POSITIVE LOGITS
    Literary
    0.91
    yarnpkg
    0.85
     quedando
    0.79
     Lynne
    0.79
    верка
    0.77
    灸
    0.77
    хой
    0.76
     Unable
    0.76
    Fig
    0.75
    を超
    0.75
    Activations Density 0.002%

    No Known Activations