© 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. 4084
    Prev
    Next
    INDEX
    Explanations

    Master, Able (titles, names)

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

    The neuron strongly activates on capitalized named entities—proper names of people, places, tribes, titles, organizations, 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_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
     relapse
    0.72
     bấm
    0.67
     torsional
    0.66
    <unused2156>
    0.65
     विविध
    0.63
     ị
    0.63
     вслед
    0.61
     YoY
    0.61
     efficacy
    0.61
    ό
    0.61
    POSITIVE LOGITS
     bernama
    0.96
     sogenannten
    0.89
     encargado
    0.89
    统治
    0.84
     llamados
    0.84
     chefe
    0.84
     llamado
    0.84
     miasta
    0.83
     chieft
    0.80
     jefe
    0.79
    Activations Density 0.292%

    No Known Activations