© 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. 15026
    Prev
    Next
    INDEX
    Explanations

    hence the following

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

    The neuron fires on bits of personal commentary and emphatic discourse (e.g. “look,” “but,” “I would,” “Jeez,” ellipses), i.e. self-referential asides and interjections.

    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
    finally
    -0.90
    首先
    -0.85
     GOOD
    -0.81
     idiots
    -0.80
    likely
    -0.80
    他们
    -0.79
     GREAT
    -0.79
     
    -0.79
    Great
    -0.78
     Så
    -0.78
    POSITIVE LOGITS
     hence
    1.18
     my
    1.17
     appunto
    1.13
     Hence
    0.98
     that
    0.97
     mentioned
    0.95
     stated
    0.94
     Именно
    0.94
     inš
    0.93
    qid
    0.93
    Activations Density 0.015%

    No Known Activations