© 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. 22806
    Prev
    Next
    INDEX
    Explanations

    more than expected, double, five

    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
    igos
    -0.10
    èĬ«
    -0.09
    ĥģ
    -0.09
    rim
    -0.09
    æĹłæĥħ
    -0.08
    oland
    -0.08
     Olsen
    -0.08
    amy
    -0.08
    榷
    -0.08
     اÙĦÙģت
    -0.08
    POSITIVE LOGITS
    丢äºĨ
    0.11
    èĩªå·±åĸľæ¬¢
    0.10
    IZER
    0.09
    /op
    0.09
    itten
    0.09
    stm
    0.09
    å®īæ£Ģ
    0.09
     baÅŁÄ±nda
    0.09
    æĿī
    0.09
    èĥ½æ»¡è¶³
    0.09
    Activations Density 0.026%

    No Known Activations