© 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. 128394
    Prev
    Next
    INDEX
    Explanations

    This neuron activates on tokens in the initial main-content sections (the first couple of numbered or titled sections) of a structured document.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    cannot do or is insufficient

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    তি
    -0.99
    볍
    -0.91
    lasz
    -0.88
     přímo
    -0.87
     necesidad
    -0.87
    Herzlich
    -0.87
    もしか
    -0.84
    slightly
    -0.84
    Noice
    -0.84
     linguri
    -0.84
    POSITIVE LOGITS
     too
    1.28
     слишком
    1.23
     enough
    1.19
     sufficiently
    1.16
     cannot
    1.16
     unable
    1.14
     sufficient
    1.12
    Unable
    1.09
     unfit
    1.00
     невозможно
    0.98
    Activations Density 0.087%

    No Known Activations