© 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. 10-GEMMASCOPE-RES-131K
    7. 15151
    Prev
    Next
    INDEX
    Explanations

    past tense verbs followed by descriptions

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

    The neuron strongly activates on past-tense or perfect‐aspect verb forms (e.g. “were,” “was,” “have,” “been,” “taken,” “had”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     currently
    -1.56
     later
    -1.40
     then
    -1.38
    currently
    -1.28
     latter
    -1.24
     ultimately
    -1.23
     finally
    -1.21
    šķ
    -1.19
     eventually
    -1.18
    нець
    -1.15
    POSITIVE LOGITS
     knew
    1.27
     nuestro
    1.27
    orre
    1.20
    тельностью
    1.19
     its
    1.16
     who
    1.11
    何か
    1.11
    させていただきました
    1.10
     серьез
    1.09
     zamiast
    1.09
    Activations Density 0.042%

    No Known Activations