© 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. 13945
    Prev
    Next
    INDEX
    Explanations

    starting `*`

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

    The neuron fires strongly on tokens that mark the start of a new section or heading (e.g. titles or the first word in a fresh article segment).

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

    periods that end sentences in formal or informational text.

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

    mentions of named-entity–rich, expository information such as organizations, locations, and formal descriptive statements.

    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
    y
    0.75
     prove
    0.72
    ʋ
    0.71
     retract
    0.70
    FromArgb
    0.68
     distinguish
    0.67
     covid
    0.67
    badan
    0.66
    檎
    0.66
     ban
    0.65
    POSITIVE LOGITS
     CLIENTI
    0.80
     overhear
    0.75
     Hfn
    0.75
     DISCUSSION
    0.73
     WITHIN
    0.73
     operazioni
    0.71
    irrah
    0.71
     ACCOUNT
    0.71
    HON
    0.71
    웨어
    0.71
    Activations Density 0.000%

    No Known Activations