© 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. 36538
    Prev
    Next
    INDEX
    Explanations

    visual cortex segmentation

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

    The neuron is a general “technical‐term” detector—it responds strongly to domain-specific content words (especially scientific/technical nouns and key terminology).

    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
     mortos
    -0.82
    iterranean
    -0.81
    न्ध
    -0.78
     tilby
    -0.77
    eloma
    -0.77
    بری
    -0.76
     stade
    -0.76
     markiert
    -0.75
    archiv
    -0.75
    чина
    -0.75
    POSITIVE LOGITS
     downstream
    0.85
     pasir
    0.82
     requ
    0.82
     without
    0.81
     flanked
    0.80
     demande
    0.80
     受
    0.78
    概要
    0.76
    ringen
    0.75
    ிறது
    0.75
    Activations Density 0.178%

    No Known Activations