© 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. 10-GEMMASCOPE-RES-131K
    7. 102797
    Prev
    Next
    INDEX
    Explanations

    album titles

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

    The neuron fires on numeric tokens—especially multi-digit figures such as sales numbers, chart positions, dates, or other statistics.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     who
    -1.51
     ventilación
    -1.18
     통해
    -1.14
     ventaja
    -1.13
    های
    -1.12
     them
    -1.12
     práci
    -1.11
    ÁCT
    -1.11
     intervento
    -1.11
    TIS
    -1.10
    POSITIVE LOGITS
    了些
    1.17
     יהיה
    1.09
     некоторы
    1.08
    OLOGIA
    1.05
    indrical
    1.05
     Könige
    1.04
     your
    1.03
     т
    1.02
     источников
    1.02
     planeta
    1.02
    Activations Density 0.042%

    No Known Activations