© 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. 48512
    Prev
    Next
    INDEX
    Explanations

    incense, incipient, incursion, incumbent

    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
    ãĥ¼ãĤº
    -0.10
    沧å·ŀ
    -0.09
    warts
    -0.09
    åĪĨå¸ĥåľ¨
    -0.09
    åĪĨå·¥
    -0.09
    æijĬ
    -0.09
    gons
    -0.09
    wart
    -0.08
    æĬ¢åħĪ
    -0.08
    å®ļä½į
    -0.08
    POSITIVE LOGITS
    ision
    0.11
    ognito
    0.10
     inc
    0.10
    idence
    0.10
    eral
    0.09
    ä¹İ
    0.09
    ional
    0.09
    -inc
    0.09
    ledon
    0.09
    ipient
    0.09
    Activations Density 0.026%

    No Known Activations