© 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. 72502
    Prev
    Next
    INDEX
    Explanations

    looking back at details

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

    This neuron detects the components of the idiomatic phrase “come to think of it” (and close variants).

    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
     in
    -2.00
     is
    -1.87
     to
    -1.44
     the
    -1.30
     pentru
    -1.23
    )
    -1.20
     when
    -1.19
    要知道
    -1.19
    issances
    -1.16
    幸好
    -1.16
    POSITIVE LOGITS
     resemble
    1.44
     empie
    1.41
    ！");
    1.36
    vänt
    1.36
     jardins
    1.34
    nehå
    1.33
     dinámico
    1.32
    初始
    1.31
     roues
    1.31
     resembled
    1.28
    Activations Density 0.007%

    No Known Activations