© 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. 65240
    Prev
    Next
    INDEX
    Explanations

    lists, descriptions, Q&A

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

    This neuron detects table column separator markers (the “|” tokens) used in table-format layouts.

    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
    Aç
    -1.26
    iree
    -1.23
    Suka
    -1.21
     increa
    -1.20
     devront
    -1.19
    Plaza
    -1.18
     dezen
    -1.17
     تريد
    -1.16
    Dimensi
    -1.16
    maar
    -1.15
    POSITIVE LOGITS
     location
    1.66
     used
    1.45
     time
    1.41
    L
    1.38
     type
    1.38
    A
    1.37
    S
    1.36
     title
    1.30
    P
    1.30
    h
    1.28
    Activations Density 0.010%

    No Known Activations