© 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. 72989
    Prev
    Next
    INDEX
    Explanations

    when we/I/you

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

    This neuron is especially sensitive to subordinating conjunctions and temporal markers (e.g. “when,” “before,” “once”) that introduce clauses.

    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
    DebuggerNonUser
    -1.42
    ۉ
    -1.40
     Diakses
    -1.29
    8
    -1.28
    一边
    -1.24
    According
    -1.24
    Busch
    -1.23
     use
    -1.23
    Aktualisiert
    -1.22
    tetlen
    -1.22
    POSITIVE LOGITS
     my
    1.74
     \
    1.72
     motoru
    1.67
    seperti
    1.55
    acterísticas
    1.53
     katk
    1.45
    đại
    1.44
    Breite
    1.44
    Karten
    1.44
     öde
    1.43
    Activations Density 0.034%

    No Known Activations