© 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. 10-GEMMASCOPE-RES-131K
    7. 74457
    Prev
    Next
    INDEX
    Explanations

    asking about what you want

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

    This neuron detects the second-person lead‐in phrase “So you want to …” (often the start of a rhetorical question to the reader).

    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
     tejto
    -1.40
    velvet
    -1.32
     châssis
    -1.28
     oignons
    -1.24
     ventilación
    -1.24
     něj
    -1.24
    ఝ
    -1.20
     caminhada
    -1.19
    ?????
    -1.18
    leds
    -1.18
    POSITIVE LOGITS
     then
    1.29
     wants
    1.23
     want
    1.20
    ;
    1.20
    それとも
    1.18
     downsides
    1.16
     Noyes
    1.16
    lograph
    1.16
    ?
    1.15
    riad
    1.14
    Activations Density 0.026%

    No Known Activations