© 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. 77734
    Prev
    Next
    INDEX
    Explanations

    method and modification

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

    The neuron activates on mentions of specific named procedures, rules, analyses, sections, or methods (e.g. “trapezoid rule,” “intent-to-treat analysis,” “Rule 405,” 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
    öhe
    -0.76
    läufig
    -0.74
    tingly
    -0.73
    噸
    -0.72
    inesi
    -0.72
     запла
    -0.72
     Kiss
    -0.71
     зака
    -0.71
    setViewportView
    -0.69
    淩
    -0.68
    POSITIVE LOGITS
     method
    1.08
    Consistency
    0.95
     methods
    0.95
     Method
    0.93
    METHOD
    0.92
     adopted
    0.91
     Modified
    0.91
     modified
    0.91
    method
    0.90
     consistent
    0.86
    Activations Density 0.016%

    No Known Activations