© 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. 10-GEMMASCOPE-RES-131K
    7. 123628
    Prev
    Next
    INDEX
    Explanations

    retrospective research types

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

    The neuron activates strongly on mentions of “retrospective” (as in “retrospective study/cohort/review”), signaling that it’s identifying references to retrospective study designs.

    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
    -
    -3.00
    </i>
    -2.83
     [
    -2.64
     we
    -2.59
    ↵↵
    -2.55
    欣賞
    -2.42
    }$
    -2.42
    </b>
    -2.42
     The
    -2.41
    ра
    -2.39
    POSITIVE LOGITS
    ୌ
    3.17
    ，“
    3.06
    酲
    3.00
    ”،
    2.98
    汭
    2.84
    哒
    2.73
    ↵
    2.70
    ⢓
    2.69
    珰
    2.67
    吖
    2.59
    Activations Density 0.016%

    No Known Activations