© 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. 57764
    Prev
    Next
    INDEX
    Explanations

    lost, loss, or loss of function/value

    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
    åĿŀ
    -0.09
    æ®ĭçķĻ
    -0.09
    oodle
    -0.08
    æķĿ
    -0.08
    è´ŁåĢº
    -0.08
    ephy
    -0.08
    å°¤åħ¶æĺ¯åľ¨
    -0.08
    ãģ¨åħ±
    -0.08
    çļĦ人çĶŁ
    -0.08
     обÑıз
    -0.08
    POSITIVE LOGITS
     lost
    0.15
    æįŁå¤±
    0.15
    ç»ıæµİæįŁå¤±
    0.14
    (loss
    0.13
    丢失
    0.13
     Lost
    0.13
    Lost
    0.12
     loss
    0.12
    丧失
    0.12
    å·¥ä½ľå²Ĺä½į
    0.11
    Activations Density 0.046%

    No Known Activations