© 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. 32209
    Prev
    Next
    INDEX
    Explanations

    architectural roles and names

    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
    æѤåĪ»
    -0.11
    ist
    -0.10
    é¡·
    -0.10
    å¤
    -0.09
    agement
    -0.09
    æĻĸ
    -0.09
     Ring
    -0.09
     Ske
    -0.09
    bourg
    -0.08
    æ°¨éħ¸
    -0.08
    POSITIVE LOGITS
    ural
    0.15
    urally
    0.13
    uate
    0.11
    ening
    0.11
    ives
    0.10
    ãģ¦ãĤĤ
    0.10
    sWith
    0.10
    ened
    0.10
    ingen
    0.10
    ue
    0.10
    Activations Density 0.006%

    No Known Activations