© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 6509
    Prev
    Next
    INDEX
    Explanations

    final step

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

    The neuron is acting as a coarse token‐position detector, firing more strongly on tokens that occur later (deeper) in the document.

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

    legal or formal language in terms and conditions, agreements, or official documents.

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

    formal legal and contractual language, especially clauses about jurisdiction, liability, prohibitions, and agreeing to terms

    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
    ева
    0.78
    jupyter
    0.76
     tokamak
    0.76
    сійскай
    0.74
    ل
    0.74
    連携
    0.73
    liferation
    0.73
    accompan
    0.71
    furan
    0.71
    кт
    0.70
    POSITIVE LOGITS
     écrire
    0.89
     posters
    0.87
     tacos
    0.86
     nije
    0.85
     copywriting
    0.83
     Coachella
    0.83
     cosplay
    0.80
     acheter
    0.80
     Iowa
    0.79
     stickers
    0.79
    Activations Density 0.001%

    No Known Activations