© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 38254
    Prev
    Next
    INDEX
    Explanations

    scientific studies and methods

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

    introductory sections of academic research papers.

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

    The neuron consistently fires on the sentence‐initial “study” or methodological lead-ins (e.g. “In the present study,” “In this study,” “Here, we …”) that introduce research methods or aims in scientific texts.

    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
     otros
    -1.15
     reisen
    -1.09
     altres
    -1.07
     també
    -1.05
     verste
    -1.05
     otra
    -1.02
     επίσης
    -1.02
     *}$
    -1.00
     también
    -0.99
     Ikke
    -0.98
    POSITIVE LOGITS
     rodríguez
    1.14
     påvir
    1.09
     médicaux
    1.06
     actuels
    1.03
     δικ
    1.01
     lópez
    1.00
     fikk
    1.00
     pelear
    1.00
     voordeel
    0.99
     veno
    0.98
    Activations Density 0.110%

    No Known Activations