© 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. 99912
    Prev
    Next
    INDEX
    Explanations

    describing qualities and states

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

    The neuron fires on descriptive, evaluative adjectives that characterize quality or intensity (e.g. “quick,” “dangerous,” “cheerful,” “concise”).

    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
     or
    -1.96
    }
    -1.70
    ),
    -1.61
     also
    -1.61
     in
    -1.59
     get
    -1.48
     For
    -1.45
     by
    -1.44
     being
    -1.41
     last
    -1.38
    POSITIVE LOGITS
     objectif
    1.75
     soldat
    1.63
     assister
    1.59
     officier
    1.59
     peint
    1.55
     condam
    1.54
     vitamine
    1.53
     cordes
    1.53
     parcs
    1.49
     applau
    1.47
    Activations Density 0.054%

    No Known Activations