© 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. 1648
    Prev
    Next
    INDEX
    Explanations

    buying, fooled, illusion, tricks

    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
    çľł
    -0.09
    åĩĨ
    -0.08
     mor
    -0.08
     outbreak
    -0.08
    ISK
    -0.08
    hatt
    -0.08
    itte
    -0.08
     pent
    -0.08
    isk
    -0.08
    åĽ
    -0.08
    POSITIVE LOGITS
     tactic
    0.11
    éªĹ
    0.10
    ä¼İ
    0.10
    è°İè¨Ģ
    0.10
    çŀĴ
    0.10
    è¯Ĩ
    0.10
    è¯ĨåĪ«
    0.10
     fooled
    0.10
     tricks
    0.10
    伪éĢł
    0.10
    Activations Density 0.047%

    No Known Activations