© 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. 68057
    Prev
    Next
    INDEX
    Explanations

    apologies and expressions of newness

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

    The neuron fires on meta‐commentary by the author—especially hedges, apologies, or first‐post/“hope I don’t screw this up” style remarks signaling uncertainty or inexperience.

    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
     we
    -1.14
     how
    -1.11
     then
    -1.09
    ֻ
    -1.01
     diverso
    -0.96
     We
    -0.96
     السياس
    -0.96
     things
    -0.95
     secretions
    -0.94
     that
    -0.93
    POSITIVE LOGITS
    皆さん
    1.17
     my
    1.09
    军团
    1.08
     frumo
    1.08
    Sharing
    1.07
     tuturor
    1.05
     oamen
    1.04
    diese
    1.04
     знают
    1.00
    ślę
    1.00
    Activations Density 0.041%

    No Known Activations