© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 38748
    Prev
    Next
    INDEX
    Explanations

    seeing how, as, that

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

    The neuron detects uses of “as” (often in multiword discourse markers like “seeing as,” “inasmuch as,” “seeing how they were…”) that introduce explanations or causal 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
    。
    -1.72
     desnuda
    -1.62
    ７
    -1.56
     artísticas
    -1.54
     eleven
    -1.53
    际
    -1.52
    時の
    -1.51
     retratos
    -1.51
    LÄ
    -1.50
     pentru
    -1.49
    POSITIVE LOGITS
    at
    1.96
     who
    1.93
     that
    1.84
     ſont
    1.61
     he
    1.61
    being
    1.60
     affront
    1.52
     they
    1.52
     continually
    1.52
     largely
    1.51
    Activations Density 0.006%

    No Known Activations