© 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. 61514
    Prev
    Next
    INDEX
    Explanations

    useless or good

    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
    çī¢åĽº
    -0.09
    æŃ£å½ĵ
    -0.09
    è°£
    -0.09
    idge
    -0.08
    ãģ¯ãģļ
    -0.08
    pler
    -0.08
    agit
    -0.08
    èĩªä»İ
    -0.08
     thiá»ĩn
    -0.08
    ä¸įçĶ±
    -0.08
    POSITIVE LOGITS
    æľīçĶ¨
    0.17
    æľīçĶ¨çļĦ
    0.17
     useless
    0.15
     useful
    0.13
     Useful
    0.13
     unus
    0.12
     usefulness
    0.12
    æľīæĦıä¹ī
    0.11
     worthless
    0.11
     meaningless
    0.11
    Activations Density 0.064%

    No Known Activations