© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 70153
    Prev
    Next
    INDEX
    Explanations

    proline allele, bonds, residue, mutations

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

    The neuron identifies scientific or medical terms, especially amino acids like "proline" and database names like "Medline".

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

    The neuron flags occurrences of the token “line,” e.g. the “–line” suffix.

    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
    s
    -2.75
     —
    -2.70
     to
    -2.52
    来た
    -2.41
     you
    -2.34
     Another
    -2.31
    ↵
    -2.27
     .
    -2.23
     o
    -2.20
    2
    -2.19
    POSITIVE LOGITS
    erintah
    3.22
    澌
    3.09
    оргаш
    2.89
    鰨
    2.70
     departamentos
    2.64
    漚
    2.63
     kabát
    2.61
    Ꮂ
    2.59
     parfü
    2.56
    神
    2.56
    Activations Density 0.015%

    No Known Activations