© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32592
    Prev
    Next
    INDEX
    Explanations

    names or specific entities followed by descriptions

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

    The neuron is tuned to pick out salient content words—particularly verbs and nouns that mark key procedural steps, outcomes, or errors in a technical narrative.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    erà
    -1.05
     trám
    -1.05
     lauk
    -1.05
     différentes
    -1.04
     итоге
    -1.02
     différents
    -1.02
     rayures
    -1.02
     tzw
    -1.01
     !!!!
    -1.01
    Congrats
    -1.01
    POSITIVE LOGITS
     bereits
    1.14
     יְ
    1.09
    información
    1.08
     Perhaps
    0.95
     Although
    0.95
    geral
    0.93
    ufficio
    0.92
    着她
    0.91
    热点
    0.89
    Here
    0.88
    Activations Density 0.092%

    No Known Activations