© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 15905
    Prev
    Next
    INDEX
    Explanations

    ChadGPT, Zorp, processing

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

    This neuron flags content-bearing (nouns, verbs, adjectives) tokens in the main argumentative or emphatic parts of a rant.

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

    casual, conversational, or sarcastic text with a personal or opinionated tone.

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

    promotional or persuasive, announcement-style phrasing that addresses or urges the audience to act.

    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_24_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
     vignette
    0.68
     collaboration
    0.67
     defining
    0.66
     collaborations
    0.65
     frustrating
    0.64
     apprenticeship
    0.63
     longitudinal
    0.63
     tena
    0.63
     နဲ့
    0.63
     mobilizing
    0.63
    POSITIVE LOGITS
    ござい
    0.74
    씐
    0.70
    icarbon
    0.69
    roben
    0.69
    Comedy
    0.68
    hehe
    0.68
    듕
    0.67
    IRCONIUM
    0.66
    ****",
    0.65
    jQuery
    0.65
    Activations Density 0.397%

    No Known Activations