© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14938
    Prev
    Next
    INDEX
    Explanations

    chess board lowercase pieces

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

    This neuron is essentially picking out unusually rare or low‐frequency tokens (i.e. those near the “end” of the model’s vocabulary), regardless of their meaning.

    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
    s
    0.96
    m
    0.88
    d
    0.79
    t
    0.77
    explore
    0.74
    motiv
    0.73
     pund
    0.72
    l
    0.71
    futbol
    0.71
     oblivion
    0.71
    POSITIVE LOGITS
     Opt
    0.86
     КП
    0.86
    Opt
    0.84
     Ako
    0.84
     Биз
    0.83
     ДО
    0.82
     ऑप्ट
    0.81
     क्रा
    0.80
    Aby
    0.80
     Меди
    0.80
    Activations Density 0.001%

    No Known Activations