© 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. 10-GEMMASCOPE-RES-131K
    7. 128834
    Prev
    Next
    INDEX
    Explanations

    supporting and connecting components

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

    The neuron strongly activates on present-participle verbs (–ing forms) that describe a device’s or material’s function or process (e.g. supporting, sealing, transporting, containing) in technical definitions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     wiederum
    -1.33
     to
    -1.33
    -1.21
    to
    -1.15
    ?
    -1.13
    ...
    -1.09
     it
    -1.09
     can
    -1.05
     denselben
    -1.05
     
    -1.04
    POSITIVE LOGITS
    ועל
    1.38
     cejas
    1.30
    AutoModerator
    1.29
     koneksi
    1.28
     TAMBÉM
    1.22
     with
    1.21
    kadang
    1.20
     besoin
    1.19
    undur
    1.19
    Parfait
    1.16
    Activations Density 0.973%
    No Known Activations