© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 3904
    Prev
    Next
    INDEX
    Explanations

    positive qualifiers and challenges

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

    The neuron lights up on uncommon content words—i.e. non-function, moderately rare or descriptive vocabulary terms.

    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
     נ
    0.50
     י
    0.48
     l
    0.46
     ৫৮
    0.46
     יי
    0.45
    0.44
     dịp
    0.44
     nhật
    0.43
     ৭০
    0.42
     concealment
    0.41
    POSITIVE LOGITS
    But
    0.64
     But
    0.55
    但是
    0.52
    yticks
    0.51
    Because
    0.50
    Amino
    0.50
    Overall
    0.50
    ToDelete
    0.50
    吡
    0.50
     પરંતુ
    0.50
    Activations Density 0.028%

    No Known Activations