© 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. 34461
    Prev
    Next
    INDEX
    Explanations

    thing is

    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
    edm
    -0.08
    御
    -0.08
    DFS
    -0.08
    å¿į
    -0.08
     pit
    -0.08
    ç¿»
    -0.08
    OWN
    -0.08
    Flying
    -0.08
    cart
    -0.08
     distractions
    -0.08
    POSITIVE LOGITS
    éĹ®é¢ĺæĺ¯
    0.13
    çļĦæĺ¯
    0.11
     thing
    0.10
    ä¸Ģ个éĹ®é¢ĺ
    0.10
    çī¹çĤ¹æĺ¯
    0.09
    éŃĦ
    0.09
    æĺ¼å¤ľ
    0.09
    thing
    0.09
    ä¸įæĪIJ
    0.09
    è¿ĻäºĽäºº
    0.09
    Activations Density 0.075%

    No Known Activations