© 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. 31-GEMMASCOPE-2-RES-16K
    7. 12606
    Prev
    Next
    INDEX
    Explanations

    pricing and sales information

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

    The neuron is tuned to pick up on numeric expressions (digits or numbers) in the text, such as section or list numbers, years, prices, dosages, etc.

    oai_token-act-pair · o4-miniTriggered 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
    devDependencies
    0.59
    ================
    0.57
    ttps
    0.55
    orges
    0.53
    !!!!!!!!!!!!!!!!
    0.53
    ----------------
    0.52
    ============
    0.52
    addOn
    0.52
    ><
    0.51
    itti
    0.51
    POSITIVE LOGITS
     सस्ते
    0.64
    安い
    0.62
     Discount
    0.61
    高い
    0.61
     деше
    0.61
     cheap
    0.60
    低
    0.59
     cheapest
    0.59
     उच्चतम
    0.59
     günstig
    0.58
    Activations Density 0.337%

    No Known Activations