© 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. 33482
    Prev
    Next
    INDEX
    Explanations

    scientific and technical vocabulary

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

    technical terms related to computing, software, and digital technologies.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron selectively activates on proper names and technical jargon—i.e. title-cased brand, product, publication, or specialized terminology tokens.

    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
     réinitial
    -1.05
    ického
    -0.98
     Füßen
    -0.97
     déchir
    -0.96
     rurales
    -0.94
    Dış
    -0.94
     étroite
    -0.93
     meste
    -0.93
    illeur
    -0.92
     trucos
    -0.90
    POSITIVE LOGITS
     allem
    0.92
     lavender
    0.91
     and
    0.90
    ۢ
    0.89
     slår
    0.88
     only
    0.85
    хъ
    0.85
     joined
    0.85
    詫
    0.85
     borough
    0.84
    Activations Density 0.081%

    No Known Activations