© 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. 24-GEMMASCOPE-2-RES-16K
    7. 9810
    Prev
    Next
    INDEX
    Explanations

    use it well

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

    The neuron is primarily detecting multi-digit numerical tokens (especially four-digit numbers, like years or large numeric values).

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

    numerical data, particularly in formatted lists, citations, and technical specifications.

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

    mentions of numbers and quantitative details (years, volumes/issues, sizes, weights, measurements) in technical or citation-style contexts.

    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
    ement
    0.79
    ać
    0.75
    aal
    0.75
    ੀ
    0.70
    asana
    0.68
     бойынша
    0.67
    ien
    0.65
    aque
    0.65
    ium
    0.64
    aan
    0.64
    POSITIVE LOGITS
    Giải
    0.79
     Unsupported
    0.78
     Buddhism
    0.75
     MST
    0.75
     findAll
    0.73
    ሽ
    0.73
     Chúa
    0.72
     Cosmo
    0.71
    욜
    0.71
     océano
    0.71
    Activations Density 0.000%

    No Known Activations