© 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. 15326
    Prev
    Next
    INDEX
    Explanations

    different unique variable

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

    This neuron activates on content-bearing words in formal or technical exposition—highlighting nouns and verbs that carry the informative substance of an explanation.

    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
    ？！”
    -1.37
     पनि
    -1.26
     אינו
    -1.25
    溼
    -1.20
     макия
    -1.18
     formule
    -1.17
     similarly
    -1.16
    জি
    -1.16
     "'",
    -1.14
    検証
    -1.14
    POSITIVE LOGITS
     then
    1.48
     menceritakan
    1.36
     allerlei
    1.36
     Então
    1.30
     rápidamente
    1.30
     berbagai
    1.30
     we
    1.29
     mérit
    1.24
     elegante
    1.24
    然后
    1.24
    Activations Density 0.027%

    No Known Activations