© 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. 25994
    Prev
    Next
    INDEX
    Explanations

    trailing characters or slashes

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

    This neuron responds to mentions of “trailing” (especially in the phrase “trailing characters” in error messages).

    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
     dê
    -0.95
     따라
    -0.93
     primeiros
    -0.88
     tô
    -0.85
    lización
    -0.85
     ordem
    -0.84
     poprzed
    -0.83
     direkte
    -0.82
     conséquence
    -0.82
     ruedas
    -0.82
    POSITIVE LOGITS
     finish
    1.20
     finishing
    1.15
     remaining
    1.13
     finished
    1.09
     any
    1.07
     Finish
    1.02
    最后
    1.02
     Finished
    1.01
     leftover
    1.00
    finalize
    0.98
    Activations Density 0.002%

    No Known Activations