© 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. 22983
    Prev
    Next
    INDEX
    Explanations

    software namespaces and technical terms

    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
    éŨå¤ĸ
    -0.10
    æľ«
    -0.09
    Grace
    -0.08
    è¾Ħ
    -0.08
    èĪ±
    -0.08
     Grace
    -0.08
     íĮIJëĭ¨
    -0.08
    éķĩæĶ¿åºľ
    -0.08
    æļĤç¼ĵ
    -0.08
    å½ĵæŶçļĦ
    -0.08
    POSITIVE LOGITS
    æ¢Ĺ
    0.09
    aries
    0.08
    éĥ¯
    0.08
    chia
    0.08
    arding
    0.08
    anners
    0.08
    çķĮéĻIJ
    0.08
    ripe
    0.08
    åıĭè°Ĭ
    0.08
    lesh
    0.08
    Activations Density 0.011%

    No Known Activations