© 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. 25525
    Prev
    Next
    INDEX
    Explanations

    names and email addresses

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

    The neuron fires on version‐control keyword expansions (e.g. $Id$, $Exp$) and especially the embedded committer usernames in source‐code comments.

    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
    があります
    -1.56
    立場
    -1.51
    と思ったら
    -1.46
    でしょうね
    -1.41
    けどね
    -1.37
    がありますが
    -1.33
    
    -1.30
    って思
    -1.30
    じゃないですか
    -1.29
    っけ
    -1.28
    POSITIVE LOGITS
    になれば
    1.52
     ТАК
    1.46
    になるので
    1.35
     Therefore
    1.34
    方も
    1.33
     OBJET
    1.33
    нівер
    1.32
    こともあり
    1.31
     БО
    1.31
     начинают
    1.28
    Activations Density 0.058%

    No Known Activations