© 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. 10-GEMMASCOPE-RES-131K
    7. 69087
    Prev
    Next
    INDEX
    Explanations

    so provides explanation

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

    The neuron detects high-level discourse markers and turn-taking cues (e.g. “So,” “Here,” “But what,” “tell us”) that introduce or transition between speaker segments.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     länder
    -1.42
     kalender
    -1.37
     FÜR
    -1.35
    퐿
    -1.34
     meil
    -1.34
     quelquefois
    -1.34
     parfois
    -1.34
    ᯅ
    -1.34
    ẕ
    -1.31
     rése
    -1.30
    POSITIVE LOGITS
     wanted
    1.63
     [
    1.49
     against
    1.43
     it
    1.42
     going
    1.41
     understand
    1.35
     halben
    1.30
     getting
    1.28
     affecting
    1.28
     move
    1.27
    Activations Density 0.016%

    No Known Activations