© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 71376
    Prev
    Next
    INDEX
    Explanations

    hesitation or transition word

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

    This neuron detects discourse-level hesitation or transition markers, such as ellipsis sequences (“…”, “….”) and filler words like “well.”

    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
     however
    -1.06
    </u>
    -1.05
    }$,
    -1.05
    当たり前
    -0.98
    kenal
    -0.97
    艹
    -0.97
     of
    -0.96
    </h5>
    -0.95
    ológia
    -0.94
    mucha
    -0.93
    POSITIVE LOGITS
     ...
    1.05
    ...
    1.05
     sprechen
    1.01
     zug
    0.96
     charred
    0.91
    ゞ
    0.91
    //!
    0.90
    Ű
    0.90
     can
    0.90
     communis
    0.89
    Activations Density 0.017%

    No Known Activations