© 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. 14053
    Prev
    Next
    INDEX
    Explanations

    words connecting clauses

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

    This neuron specializes in picking out formal technical–expository signal words and connectors (e.g. thereby, objective, to, coordination) that structure patent‐style or scientific explanations.

    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
    娌
    -1.02
     devront
    -1.00
     chré
    -1.00
     görmek
    -0.95
     fizeram
    -0.94
    ユニクロ
    -0.93
    ナイキ
    -0.93
     Especially
    -0.93
    を作り
    -0.92
    そして
    -0.91
    POSITIVE LOGITS
    È
    1.13
     pursuant
    1.06
    Α
    1.02
    B
    1.01
     XI
    1.00
    T
    0.97
    PR
    0.97
     निम्नलिखित
    0.96
    ?!
    0.94
    жо
    0.94
    Activations Density 0.582%

    No Known Activations