© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 10150
    Prev
    Next
    INDEX
    Explanations

    discussion within

    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
    ä»»ä½ķ人éĥ½
    -0.09
    ä½İå»ī
    -0.09
    çľīæ¯Ľ
    -0.08
    éĽĨè£ħ
    -0.08
    \↵
    -0.08
     absentee
    -0.08
    ADI
    -0.08
     açıkl
    -0.08
    æłĩåĩĨåĮĸ
    -0.08
    ffer
    -0.08
    POSITIVE LOGITS
    åĪĨä¼ļ
    0.09
    _STS
    0.09
    bsites
    0.09
    æĤłæĤł
    0.08
    çļĦåĬĽéĩı
    0.08
    æĹĭ
    0.08
    stamp
    0.08
    æ·±åħ¥
    0.08
    åĪĨæĪIJ
    0.08
    aser
    0.08
    Activations Density 0.042%

    No Known Activations