© 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. 15848
    Prev
    Next
    INDEX
    Explanations

    medical and technical terms

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

    programming/coding-related terms and library references in software documentation.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on tokens that name external libraries, frameworks, or modules (e.g. Zeppelin, Module, Sherlock, enclave, xdt, sinks, upgrade).

    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
    ないです
    -1.20
     "";
    -1.04
    PropTypes
    -1.04
     Atenas
    -1.02
     toont
    -1.02
     kanilang
    -1.02
     TÉCN
    -1.00
    __':
    -1.00
    ']='
    -1.00
     morrer
    -1.00
    POSITIVE LOGITS
    ↵↵↵↵↵↵↵↵↵↵
    1.10
    1.10
     meisten
    1.09
     earthquakes
    1.08
     verschieden
    1.05
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    1.03
     häufiger
    1.02
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    1.02
     proce
    1.00
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.99
    Activations Density 0.000%
    No Known Activations