© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3394
    Prev
    Next
    INDEX
    Explanations

    This neuron flags large numeric tokens—especially multi‐digit reference numbers (e.g. journal volumes, page ranges, years) in bibliographic citations.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    the word "at" when it appears in formal contexts like academic credentials, institutional affiliations, or professional accomplishments.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    mentions of receiving awards or recognitions, often alongside formal citation-style numbers such as years or DOI-like identifiers.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     forensics
    0.73
    <h1>
    0.68
     clandest
    0.66
     RAND
    0.65
     aspects
    0.64
    --
    0.63
     positions
    0.63
    origins
    0.62
    -|
    0.61
     delegations
    0.61
    POSITIVE LOGITS
    тным
    1.09
    ʘ
    1.05
    asını
    0.94
    ებში
    0.91
     сдела
    0.90
    𝘱
    0.90
     infección
    0.87
     аллер
    0.87
    для
    0.86
     bữa
    0.86
    Activations Density 0.000%

    No Known Activations