© 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. 72904
    Prev
    Next
    INDEX
    Explanations

    names and places

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

    The neuron specializes in picking out segments of multi‐token proper names—especially surnames of players—in sports/tournament write-ups.

    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
     voulez
    -1.10
    roep
    -1.09
    Недостатки
    -1.06
     camisas
    -1.02
     когда
    -1.01
     Anfitrión
    -0.98
     عندما
    -0.96
    Either
    -0.96
     mapas
    -0.96
     Nós
    -0.94
    POSITIVE LOGITS
    ans
    1.09
    anga
    1.09
     prosegu
    1.04
    genous
    1.02
    的重要
    1.02
     from
    1.00
     וּ
    0.99
    ach
    0.98
    的主要
    0.97
     verschiedener
    0.96
    Activations Density 0.001%

    No Known Activations