© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 10081
    Prev
    Next
    INDEX
    Explanations

    separators or sentence terminators

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

    The neuron flags evaluative or judgment-laden content words—terms that convey an assessment or stance (e.g. “sure,” “potential,” “success,” “discouraging,” “technique,” etc.).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     albeit
    1.19
    albeit
    1.07
     aunque
    1.01
     although
    0.96
     though
    0.94
    虽然
    0.93
    虽
    0.89
    雖
    0.89
    aunque
    0.86
    though
    0.85
    POSITIVE LOGITS
    …,
    1.06
     ...,
    1.04
     …,
    1.03
    ...,
    1.02
     HOWEVER
    0.93
     However
    0.93
    ^{*},
    0.93
    However
    0.92
     Međutim
    0.91
    (),
    0.84
    Activations Density 0.093%

    No Known Activations