© 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. 25591
    Prev
    Next
    INDEX
    Explanations

    reciprocation, similarity, agreement

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

    The neuron selectively fires on semantically rich content words (not function words)—e.g. key nouns, verbs, pronouns, and intensifiers—rather than on common stop‐words or punctuation.

    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
     there
    -0.82
     said
    -0.81
    <em>
    -0.80
    警
    -0.80
    ते
    -0.80
     świecie
    -0.79
     hinzugefügt
    -0.79
    曹
    -0.79
    贋
    -0.78
     jsonArray
    -0.78
    POSITIVE LOGITS
    Likewise
    1.23
     тоже
    1.19
     Likewise
    1.16
     reciproc
    1.13
     likewise
    1.02
    ísima
    0.92
    ⦑
    0.89
    こちらも
    0.88
    ・・・・・
    0.87
    slika
    0.85
    Activations Density 0.018%

    No Known Activations