© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 110508
    Prev
    Next
    INDEX
    Explanations

    choose between

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

    This neuron fires strongly on the first salient content word at the start of a phrase or clause (i.e. the first non‐function word that carries the main meaning of a new segment).

    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
     vitrine
    -1.46
    ONLY
    -1.45
     faciles
    -1.40
     bené
    -1.39
     such
    -1.39
    ı
    -1.32
    ïque
    -1.32
    vidla
    -1.28
    .”
    -1.27
     abiding
    -1.27
    POSITIVE LOGITS
    9
    1.44
     possibilité
    1.29
    不錯
    1.27
     شك
    1.26
     pomocí
    1.25
    Ⓡ
    1.25
     suficiente
    1.24
    lelő
    1.23
     privileged
    1.23
     lorsque
    1.23
    Activations Density 0.021%

    No Known Activations