© 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. 25530
    Prev
    Next
    INDEX
    Explanations

    medical literature and programming

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

    technical or programming terms that appear in code, documentation, or scientific papers.

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

    This neuron responds to isolated technical tokens—especially all‐caps acronyms, proper‐name fragments, and numeric or alphanumeric labels.

    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
     after
    -1.20
     Because
    -1.18
     Despite
    -1.10
     should
    -1.05
     Since
    -1.02
     несмотря
    -1.02
     Although
    -1.02
     during
    -1.01
     makes
    -1.01
     can
    -0.98
    POSITIVE LOGITS
    ﾒ
    1.22
    嗫
    1.19
     miniaturka
    1.16
    
    1.16
     zoude
    1.16
     premiers
    1.13
    athered
    1.13
    饸
    1.12
     dezelve
    1.11
     creato
    1.10
    Activations Density 0.075%

    No Known Activations