© 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. 34-GEMMASCOPE-RES-131K
    7. 119270
    Prev
    Next
    INDEX
    Explanations

    research mechanisms and findings

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

    This neuron detects scientific and academic language in technical papers, particularly in biochemistry, medicine, and physics research articles.

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

    The neuron is detecting a token’s position in the document, producing higher activations for words that occur later in the text.

    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
     preventative
    -1.27
     правительства
    -1.22
     numerosas
    -1.09
    enares
    -1.02
     неболь
    -1.02
     ڤ
    -1.02
     salespeople
    -1.01
     konsep
    -1.00
     formulas
    -1.00
     entretenimiento
    -1.00
    POSITIVE LOGITS
     atleast
    1.18
     aforesaid
    1.17
     along
    1.07
     upto
    1.06
     improvised
    1.02
     enlisted
    1.02
     popularly
    0.97
    Earlier
    0.95
     improvis
    0.94
    engg
    0.94
    Activations Density 0.261%

    No Known Activations