© 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. 33001
    Prev
    Next
    INDEX
    Explanations

    necessity or obligation

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

    The neuron is picking out infinitival or modal‐verb constructions—especially the “to be” (or “need to be”/“must be”) pattern.

    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
    方針
    -1.43
    は
    -1.41
    <bos>
    -1.38
    調味料
    -1.34
    などが
    -1.21
    もあって
    -1.21
    が
    -1.19
    と
    -1.18
     cannot
    -1.18
     in
    -1.18
    POSITIVE LOGITS
    為に
    1.83
    ところに
    1.59
    久し
    1.57
    にほんブログ村
    1.55
    さぁ
    1.52
    おそ
    1.47
    明日は
    1.44
    えっ
    1.42
    の方に
    1.41
    マジで
    1.41
    Activations Density 0.060%

    No Known Activations