© 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. 34349
    Prev
    Next
    INDEX
    Explanations

    question, let, genetic, order

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

    large empty whitespace in documents.

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

    This neuron activates on standalone numeric tokens (multi‐digit numbers) in the text.

    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
     pensó
    -1.14
    皤
    -1.07
    𝐼
    -1.05
    Fordítás
    -1.03
     pérdidas
    -1.02
    معلومات
    -1.01
    ždy
    -1.00
     sizin
    -1.00
    Tác
    -1.00
    遊ぶ
    -0.97
    POSITIVE LOGITS
     with
    1.23
     will
    1.13
     school
    1.10
     helps
    1.07
     then
    1.07
    “
    1.02
    Cadastrar
    1.02
     Chicks
    1.00
    日至
    1.00
     takes
    0.98
    Activations Density 0.000%

    No Known Activations