© 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. 12300
    Prev
    Next
    INDEX
    Explanations

    guaranteed not

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

    The neuron fires on tokens that look like metadata or reference markers—dates, section titles, citation words, and other “label‐style” tokens.

    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
    鄔
    0.88
     artillery
    0.87
     transceiver
    0.86
     самостоятельно
    0.85
    主体
    0.84
    ţe
    0.82
    یر
    0.82
     এছাড়াও
    0.82
     Kinney
    0.82
    arthen
    0.81
    POSITIVE LOGITS
    vera
    0.71
    LED
    0.71
    Despite
    0.71
    琰
    0.68
    ],'
    0.68
    vering
    0.67
    Bayern
    0.67
    vitamin
    0.63
    nement
    0.63
    iad
    0.63
    Activations Density 0.000%

    No Known Activations