© 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. 57080
    Prev
    Next
    INDEX
    Explanations

    scientific reliability authenticity

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

    The neuron detects structural/legal document headings and numbered rules or section citations (e.g. “Rule 902 – Evidence That Is Self-Authenticating,” numbered lists, and similar formatted headings).

    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
    ```
    -0.73
    Pero
    -0.70
    ágot
    -0.70
     penal
    -0.70
     Initi
    -0.69
    arza
    -0.69
     Initiation
    -0.66
    Palla
    -0.66
    preds
    -0.66
    zení
    -0.65
    POSITIVE LOGITS
     Frye
    1.03
     científicas
    1.00
     scientific
    0.99
     reliability
    0.96
     Reliability
    0.96
     science
    0.93
    Fry
    0.85
     briefed
    0.85
     Science
    0.85
     régulièrement
    0.84
    Activations Density 0.012%

    No Known Activations