© 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. 22-GEMMASCOPE-RES-131K
    7. 16031
    Prev
    Next
    INDEX
    Explanations

    target audience and anyone

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

    This neuron detects “Who” question phrases—especially headings or titles beginning with “Who is…” or “Who needs…”

    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
     after
    -1.20
     that
    -1.15
     even
    -1.06
     on
    -1.05
     at
    -1.01
     later
    -1.00
     through
    -0.97
     additional
    -0.96
     one
    -0.96
     After
    -0.90
    POSITIVE LOGITS
     anyone
    1.55
    anyone
    1.28
    Anyone
    1.26
     ideal
    1.23
    EVERY
    1.13
     anybody
    1.12
    those
    1.09
     cualquier
    1.09
    ANY
    1.08
     trouvera
    1.07
    Activations Density 0.081%

    No Known Activations