© 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. 26205
    Prev
    Next
    INDEX
    Explanations

    numbers and codes after punctuation

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

    The neuron is strongly triggered by numeric tokens and digit sequences (e.g., dates, version numbers, rule citations).

    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
    <0xF0>
    -1.09
    velocityY
    -1.05
    lcm
    -1.02
     januari
    -1.02
     javax
    -1.01
    arbeid
    -1.00
     špec
    -1.00
    də
    -0.98
    }^{-
    -0.97
     цену
    -0.96
    POSITIVE LOGITS
     gustaba
    1.20
    灃
    1.05
    <bos>
    1.05
    Scooby
    1.03
     später
    1.03
    stays
    1.03
    Cinco
    1.02
     a
    1.02
    Shipped
    1.02
    言えば
    1.01
    Activations Density 0.017%

    No Known Activations