© 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. 31-GEMMASCOPE-2-RES-16K
    7. 401
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly responds to the start-of-sequence token and to the article “a” at the beginnings of sentences or segments.

    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_31_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
     C
    0.50
     delimiters
    0.49
    C
    0.47
     E
    0.46
     R
    0.46
     S
    0.45
     Use
    0.45
     M
    0.44
     G
    0.43
     identifiers
    0.42
    POSITIVE LOGITS
     zeer
    0.59
     bardzo
    0.59
     multimillion
    0.59
     hotly
    0.58
     veoma
    0.57
     très
    0.55
     lucrative
    0.55
     flamboyant
    0.54
     sprawling
    0.54
     veľmi
    0.54
    Activations Density 2.134%

    No Known Activations