© 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. 59687
    Prev
    Next
    INDEX
    Explanations

    multi-language documentation

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

    This neuron responds to the punctuation markers (colons, dashes, block-quote symbols) that define metadata fields and separators (e.g. in a Markdown/YAML front-matter header).

    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
     hábit
    -1.27
     You
    -1.23
     espectaculares
    -1.23
     their
    -1.20
     Auftrag
    -1.18
    partements
    -1.18
     bosques
    -1.14
     площадь
    -1.11
     Many
    -1.10
     outstanding
    -1.10
    POSITIVE LOGITS
    1.38
    espé
    1.32
     هذا
    1.28
     coś
    1.22
    wiście
    1.18
    𝕍
    1.17
    ρους
    1.17
     chỉ
    1.16
     haremos
    1.16
     they
    1.16
    Activations Density 0.081%

    No Known Activations