© 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. 22-GEMMASCOPE-RES-131K
    7. 65048
    Prev
    Next
    INDEX
    Explanations

    reporting who spoke

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

    The neuron strongly fires on bits of reported speech attribution (e.g. “Court said,” “James Jarvis said,” “the CEO said,” “she added,” 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
     tulisan
    -0.99
     escrita
    -0.96
     escrito
    -0.94
     mivel
    -0.90
     Während
    -0.84
    رایش
    -0.82
    écrire
    -0.80
    〜。
    -0.78
     thwart
    -0.77
     insulators
    -0.77
    POSITIVE LOGITS
     interview
    1.34
     interviews
    1.27
    interview
    1.25
     interviewed
    1.20
     speaking
    1.13
     Interview
    1.13
     video
    1.08
    Interview
    1.06
     said
    1.02
     interviewer
    1.00
    Activations Density 0.026%

    No Known Activations