© 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. 57500
    Prev
    Next
    INDEX
    Explanations

    permanently closing, reducing, or damaging

    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
    è¿ĺæľīä»Ģä¹Ī
    -0.10
    jax
    -0.09
    oms
    -0.09
    atched
    -0.09
    facts
    -0.09
    ðŁı»
    -0.09
    oused
    -0.09
    asjon
    -0.08
    sing
    -0.08
     CIM
    -0.08
    POSITIVE LOGITS
    æ°¸ä¹ħ
    0.10
    ely
    0.09
     permanently
    0.09
    ä¸Ģ页
    0.09
    çĹ¼
    0.09
    æĦıä¹īä¸Ĭ
    0.09
    æ°¸è¿ľ
    0.09
    æ°¸éģł
    0.09
    éĺIJéĩĬ
    0.09
    æĦıä¹īä¸ĬçļĦ
    0.09
    Activations Density 0.014%

    No Known Activations