© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 130336
    Prev
    Next
    INDEX
    Explanations

    food preparation and texture

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

    texture-related food descriptions like "crispy," "tender," "chewy," and words describing food doneness.

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

    The neuron activates on descriptive adjectives of food texture or doneness (e.g. “crisp,” “tender,” “golden”).

    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
    aimeJ
    -1.42
    rogens
    -1.35
    ebly
    -1.29
    chery
    -1.29
    锨
    -1.28
    termilk
    -1.28
    iibo
    -1.27
    opies
    -1.27
    owan
    -1.26
    弱い
    -1.26
    POSITIVE LOGITS
     all
    1.30
    Some
    1.25
    :{
    
    1.22
    because
    1.20
     my
    1.20
     assorted
    1.20
     خبرگزاری
    1.18
    bunga
    1.16
    Just
    1.15
     thoroughly
    1.14
    Activations Density 0.014%

    No Known Activations