© 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. 34997
    Prev
    Next
    INDEX
    Explanations

    what makes / what's the reason

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

    The neuron fires on explanatory or causal clauses introduced by “what” or “which” (e.g. “what makes,” “what was,” “which probably was the draw,” “this is what…”).

    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
     confirmer
    -1.11
    licitations
    -1.08
     bottes
    -1.02
    kowej
    -1.00
    📲
    -1.00
     paroles
    -0.99
     contradictions
    -0.98
    datable
    -0.96
    haled
    -0.94
    читан
    -0.93
    POSITIVE LOGITS
     why
    2.75
     reason
    2.00
     reasons
    1.82
     Why
    1.62
    why
    1.60
    Why
    1.52
     razones
    1.44
     почему
    1.41
     warum
    1.39
     причина
    1.38
    Activations Density 0.060%

    No Known Activations