© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 71858
    Prev
    Next
    INDEX
    Explanations

    competition outcomes and rankings

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

    This neuron detects numeric labels and ordinals denoting tournament stages or placement matches (e.g., “5th/6th place playoff,” “Bronze medal match,” “Quarterfinals”).

    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
    Recursos
    -1.13
     corrom
    -1.08
     konkre
    -1.07
     relatif
    -1.00
     déclen
    -0.97
    aları
    -0.95
     interdit
    -0.94
    garage
    -0.93
    Dossier
    -0.93
     humain
    -0.91
    POSITIVE LOGITS
     consolation
    1.16
     with
    1.01
    phée
    1.00
     bottom
    0.96
     after
    0.94
    હે
    0.94
     loses
    0.93
     went
    0.91
     veniva
    0.91
     loser
    0.90
    Activations Density 0.011%

    No Known Activations