© 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. 34-GEMMASCOPE-RES-131K
    7. 109364
    Prev
    Next
    INDEX
    Explanations

    fear or worry

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

    The neuron fires on very high-frequency “function” tokens—common short words (prepositions, pronouns) and punctuation—regardless of content.

    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
     vapormax
    -0.92
     basili
    -0.92
     ベクター
    -0.91
     Vek
    -0.88
     VECTOR
    -0.87
    帏
    -0.85
     Pagina
    -0.85
     minera
    -0.84
    isesta
    -0.83
     alho
    -0.83
    POSITIVE LOGITS
     fear
    6.13
     fears
    5.19
     feared
    5.13
     fearing
    5.03
    fear
    4.78
     afraid
    4.78
     Fear
    4.63
    Fear
    4.44
    怕
    3.78
     FEAR
    3.73
    Activations Density 0.111%

    No Known Activations