© 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. 13326
    Prev
    Next
    INDEX
    Explanations

    `can treat` `called directly` `is different`

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

    The neuron strongly activates on multiword terms naming organized programs or services—especially structured training or treatment offerings.

    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
    UInt
    0.75
     UInt
    0.71
    CharPtr
    0.71
     Electrode
    0.70
    ｃｍ
    0.66
    setKeepCallback
    0.65
    Ｐ
    0.64
     Έ
    0.64
     अल्को
    0.64
    
    0.64
    POSITIVE LOGITS
     eviction
    0.68
    atgu
    0.60
     layoff
    0.59
     lottery
    0.58
     ega
    0.56
     feral
    0.56
     paves
    0.56
     villa
    0.55
     bef
    0.55
     layoffs
    0.55
    Activations Density 0.007%

    No Known Activations