© 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. 9845
    Prev
    Next
    INDEX
    Explanations

    used, returned, crucial, included, covered, blocked

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

    The neuron spikes on uses of the verb “to be” (especially auxiliary “be” forms like “being,” “are,” “is”) marking passive‐voice or copular constructions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    the plural form of the verb "to be" (are/were) in passive voice constructions.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    forms of the verb “to be,” especially when used as copulas or auxiliary verbs.

    oai_token-act-pair · gpt-5Triggered 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
     perpetrated
    0.81
     inflicted
    0.76
     gerçekleş
    0.76
     alleviated
    0.75
     dilakukan
    0.74
     marred
    0.73
     mitigated
    0.72
     thwarted
    0.72
     triggered
    0.72
     occasioned
    0.72
    POSITIVE LOGITS
     hereby
    0.71
     included
    0.63
    Included
    0.61
     Included
    0.60
     অন্তর্ভুক্ত
    0.60
     வைத்திருக்க
    0.58
     joins
    0.55
     receives
    0.55
     sits
    0.55
     keeps
    0.54
    Activations Density 0.185%

    No Known Activations