© 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. 22-GEMMASCOPE-RES-131K
    7. 95453
    Prev
    Next
    INDEX
    Explanations

    permanent states or actions

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

    The neuron is detecting occurrences of the word “permanent” (in its various morphological forms).

    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
     \"
    -2.38
    蟥
    -2.23
     ``
    -2.23
     terrific
    -2.20
     magnific
    -2.20
    那我
    -2.20
     impromptu
    -2.19
     We
    -2.19
    槌
    -2.17
     当
    -2.16
    POSITIVE LOGITS
    just
    3.16
     Surprisingly
    2.84
    what
    2.78
    Although
    2.61
    Like
    2.61
     Interestingly
    2.59
    Even
    2.53
    ஂ
    2.53
     Regardless
    2.52
     Undoubtedly
    2.50
    Activations Density 0.008%

    No Known Activations