© 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-3-27B-IT
    5. 40-GEMMASCOPE-2-RES-262K
    7. 12732
    Prev
    Next
    INDEX
    Explanations

    engaged in conflict or struggle

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

    function words that serve as grammatical glue—especially prepositions and auxiliary verb contractions linking actions or clauses

    oai_token-act-pair · gpt-5Triggered by @yooniel31

    argumentative dialogue

    np_max-act · gpt-5Triggered by @yooniel31

    The neuron activates strongly on words that describe taking away or withholding possessions (e.g. “confiscation,” “footwear,” “defending,” “preoccupied”).

    oai_token-act-pair · o4-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_40_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    jml
    0.43
     OpenAI
    0.40
     projection
    0.39
    rb
    0.39
    indi
    0.38
     {:.
    0.38
     im
    0.37
    ractive
    0.37
    INSTANCE
    0.37
     câncer
    0.37
    POSITIVE LOGITS
     valiant
    0.79
     heinous
    0.78
     glorious
    0.76
     nefarious
    0.75
     vanqu
    0.73
     shenanigans
    0.72
     perilous
    0.72
     trusty
    0.71
     mighty
    0.70
     weaponry
    0.69
    Activations Density 0.059%

    No Known Activations