© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 5667
    Prev
    Next
    INDEX
    Explanations

    actions and outcomes

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

    The neuron fires on content-heavy action verbs—strongly activating on words that denote purposeful actions or processes (e.g. face, find, think, strengthen, improve).

    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_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
    ;
    0.68
    -
    0.65
     ledge
    0.57
    /
    0.56
    1
    0.54
    }
    0.53
     merely
    0.53
    ft
    0.50
     everybody
    0.50
    (
    0.49
    POSITIVE LOGITS
    
    0.68
     FGFR
    0.65
     TRPV
    0.63
     다양한
    0.63
     높은
    0.61
     arxiv
    0.61
     hemolysis
    0.61
     overclock
    0.61
     PDGF
    0.61
     adsorption
    0.60
    Activations Density 0.000%

    No Known Activations