© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 12909
    Prev
    Next
    INDEX
    Explanations

    all data

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

    The neuron detects formal legal or regulatory disclaimer phrasing (e.g. “represent and warrant that,” “hereby state that,” “fully understand,” 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_22/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
     become
    -1.11
     but
    -1.09
     automáticamente
    -1.07
    ^*$
    -1.06
     still
    -1.06
     instead
    -1.05
     Prä
    -1.04
     eventually
    -1.02
     automatically
    -1.02
     gets
    -1.01
    POSITIVE LOGITS
    😤
    1.20
     satisfactory
    1.13
    🤯
    1.10
    🥴
    1.09
     Sien
    1.09
    🙆
    1.08
    😫
    1.07
    meet
    1.05
     garantir
    1.05
     シール
    1.04
    Activations Density 0.007%

    No Known Activations