© 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. 113058
    Prev
    Next
    INDEX
    Explanations

    states or properties

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

    The neuron is looking for tokens carrying affix‐based property modifiers—especially those beginning with common scientific/English prefixes like non-, in-, re-, anti- (e.g. non-renewable, inflexible, insoluble, antisolvent).

    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
     through
    -1.63
     before
    -1.56
     wholly
    -1.44
     because
    -1.44
     wildly
    -1.28
     burgeoning
    -1.27
    翌日
    -1.27
    There
    -1.25
     allowing
    -1.25
     after
    -1.24
    POSITIVE LOGITS
     ventilación
    1.44
    weisen
    1.33
    ться
    1.32
    
    1.25
    might
    1.25
    didn
    1.25
     sellado
    1.23
     indiquer
    1.20
    arding
    1.18
     tylko
    1.17
    Activations Density 0.108%

    No Known Activations