© 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. 22-GEMMASCOPE-RES-131K
    7. 65264
    Prev
    Next
    INDEX
    Explanations

    substitute or replace

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

    The neuron detects nouns that name items being proposed as replacements or alternatives in “substitute/alternative” contexts.

    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
     inni
    -0.96
     незавершена
    -0.93
    比べると
    -0.91
    zeß
    -0.90
    ule
    -0.89
    おまけ
    -0.89
    нении
    -0.86
    kawa
    -0.84
    unov
    -0.83
     labd
    -0.82
    POSITIVE LOGITS
     miesto
    1.25
     because
    1.22
     Substitute
    1.18
     substitute
    1.13
     replace
    1.07
    utusan
    1.03
     sostitu
    1.03
     karena
    1.03
     replaced
    1.01
     replacing
    1.00
    Activations Density 0.070%

    No Known Activations