© 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. 4317
    Prev
    Next
    INDEX
    Explanations

    invalid choice or input

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

    The neuron detects error or failure keywords (e.g. “unsupported,” “invalid”) signaling an error message.

    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
    霢
    0.66
    делать
    0.61
    hect
    0.61
     ऑन
    0.59
    albeit
    0.58
    isDisabled
    0.58
    卧室
    0.57
    usuarios
    0.56
    awcy
    0.56
    шать
    0.55
    POSITIVE LOGITS
    !
    0.73
    !!
    0.70
    거나
    0.68
    !!”
    0.68
    !”
    0.65
     !!
    0.63
     или
    0.63
     atau
    0.63
    或是
    0.62
     või
    0.61
    Activations Density 0.000%

    No Known Activations