© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 36240
    Prev
    Next
    INDEX
    Explanations

    about when

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

    The neuron activates on verbs that express a personal attitude or emotional stance (e.g. want, worry, care, interested).

    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
     Those
    -1.64
     during
    -1.56
     on
    -1.55
     can
    -1.55
     before
    -1.48
     Many
    -1.48
     specified
    -1.46
    遑
    -1.45
    iapkan
    -1.42
    謊
    -1.41
    POSITIVE LOGITS
     being
    1.77
    盉
    1.75
     the
    1.52
    衚
    1.50
    轳
    1.47
    衕
    1.41
    屃
    1.38
    𓐍
    1.38
     aDecoder
    1.38
     camara
    1.37
    Activations Density 0.038%

    No Known Activations