© 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. 78312
    Prev
    Next
    INDEX
    Explanations

    technical or structured input

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

    This neuron detects numeric tokens and count-related terms—i.e. digits or words indicating quantities such as IDs, occurrence counts, layer counts, etc.

    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
     during
    -1.22
     even
    -1.02
     since
    -1.01
     while
    -0.90
     before
    -0.90
     system
    -0.89
     During
    -0.86
     remained
    -0.86
     Since
    -0.84
     in
    -0.81
    POSITIVE LOGITS
     Expectation
    0.96
    enterOuterAlt
    0.96
     recente
    0.90
    RuleContext
    0.88
     Currie
    0.86
     sensibili
    0.86
    いきなり
    0.85
     solem
    0.85
     efek
    0.85
     писто
    0.82
    Activations Density 0.000%

    No Known Activations