© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34482
    Prev
    Next
    INDEX
    Explanations

    completing phrases or names

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

    The neuron strongly activates on the main topic or headline tokens at the start of a new section or article (e.g. key words in titles or section headings).

    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
     vard
    -1.08
     Compte
    -1.07
     coq
    -1.05
     pegat
    -1.05
     nand
    -1.03
     Tierra
    -1.02
     eſſ
    -1.02
     ister
    -1.02
     rek
    -1.02
     nus
    -1.02
    POSITIVE LOGITS
     to
    1.05
     Those
    1.03
     Each
    1.02
     However
    0.98
    Those
    0.96
     has
    0.95
     }
    0.94
     from
    0.93
     at
    0.92
     not
    0.92
    Activations Density 0.138%

    No Known Activations