© 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. 34-GEMMASCOPE-RES-131K
    7. 3024
    Prev
    Next
    INDEX
    Explanations

    specialized units and placements

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

    The neuron strongly activates on technical or domain-specific keywords you’d see in formal education or medical contexts—terms like “special,” “education,” “treatment,” “rehabilitation,” “students,” etc.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    街头
    -0.82
     Affordable
    -0.73
    leg
    -0.72
    tee
    -0.71
     pessoais
    -0.71
     Rentals
    -0.70
     féd
    -0.70
    cking
    -0.70
    lei
    -0.70
     revelou
    -0.70
    POSITIVE LOGITS
     residential
    1.40
     placement
    1.34
     specialized
    1.31
     placements
    1.25
     alternative
    1.10
     Placement
    1.09
    specialized
    1.09
     unit
    1.06
     specialised
    1.06
    Specialized
    1.05
    Activations Density 0.044%

    No Known Activations