© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 56522
    Prev
    Next
    INDEX
    Explanations

    wanna/gonna/gotta followed by a verb

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    eing
    -0.10
    an
    -0.09
    è¿Ļæīį
    -0.09
    овое
    -0.08
    é¦ĸåĪĽ
    -0.08
    ç»ıæŁ¥
    -0.08
    y
    -0.08
    SIZE
    -0.08
    ening
    -0.08
    ened
    -0.08
    POSITIVE LOGITS
    è°´
    0.10
    åģļçļĦäºĭæĥħ
    0.10
    ellas
    0.10
    个å¤ļ
    0.10
    æĮĩæľĽ
    0.10
    ment
    0.10
    æĬĬè¿ĻäºĽ
    0.10
    æľī大éĩıçļĦ
    0.09
    iosa
    0.09
    ä¼°
    0.09
    Activations Density 0.007%

    No Known Activations