© 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. 10612
    Prev
    Next
    INDEX
    Explanations

    compression utilities

    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
    Argb
    -0.10
    dish
    -0.10
    ç¾İå¦Ļ
    -0.09
    /respond
    -0.09
    éĻĮ
    -0.09
    æĭįçħ§
    -0.09
    å¼ĥ
    -0.09
    æŃ¥è¡Į
    -0.09
    ç®Ĺ
    -0.09
    ÑģÑĤÑĢел
    -0.09
    POSITIVE LOGITS
    åİĭ缩
    0.14
     compression
    0.10
    orge
    0.10
    ade
    0.10
    å°ģ
    0.09
    ipping
    0.09
    iveness
    0.09
     compress
    0.09
     compressed
    0.09
    æīĵ
    0.09
    Activations Density 0.107%

    No Known Activations