© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 5700
    Prev
    Next
    INDEX
    Explanations

    role attachment snack old playing

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

    The neuron is strongly triggered by the letter‐sequence “deriv” (and its fragments like “Der,” “iv,” “ate,” etc.), i.e. parts of the word “derivative.”

    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
    )
    0.82
    ),
    0.79
     endorph
    0.77
    )...
    0.77
    )#
    0.77
    ):
    0.73
     Ape
    0.73
     Pads
    0.73
    ))
    0.71
    )....
    0.71
    POSITIVE LOGITS
    riften
    0.88
    महिला
    0.86
    citizens
    0.81
    महिलाओं
    0.80
    ്രി
    0.80
    होस्
    0.79
    हालांकि
    0.79
    titled
    0.77
    treated
    0.77
    s
    0.76
    Activations Density 0.001%

    No Known Activations