© 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. 23625
    Prev
    Next
    INDEX
    Explanations

    survey investigates

    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
    ities
    -0.11
    lessly
    -0.10
    pend
    -0.09
    æĹłæĥħ
    -0.08
    izio
    -0.08
    idl
    -0.08
    ãĤĪãģĨãģª
    -0.08
    chaft
    -0.08
    enberg
    -0.08
    ãĤ¢ãĤ¦ãĥĪ
    -0.08
    POSITIVE LOGITS
    ing
    0.13
    ors
    0.11
    ingly
    0.11
    åıĸè¯ģ
    0.10
    éĹ®åį·
    0.10
    able
    0.10
    ork
    0.09
    ING
    0.09
    ably
    0.09
    è¡·
    0.09
    Activations Density 0.025%

    No Known Activations