© 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. 49964
    Prev
    Next
    INDEX
    Explanations

    git commands and concepts

    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
    ãĤ«ãĥ«
    -0.10
    èĥĨ
    -0.09
     Sight
    -0.09
    èIJ½åľ°
    -0.09
     captivity
    -0.08
     sight
    -0.08
     nab
    -0.08
    èĨ½
    -0.08
    ä¸ĢåĪĢ
    -0.08
     sweep
    -0.08
    POSITIVE LOGITS
    æ²ı
    0.10
    æľºæ¢°è®¾å¤ĩ
    0.10
    çł¾
    0.10
    umbing
    0.10
    ëī´ìĬ¤
    0.09
    .cid
    0.09
    umph
    0.09
    IndexChanged
    0.09
    andise
    0.09
    泡
    0.09
    Activations Density 0.292%

    No Known Activations