© 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. 31318
    Prev
    Next
    INDEX
    Explanations

    tricks and tools

    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
    bare
    -0.09
    åĿ¨
    -0.09
    æ¶ij
    -0.09
    esterday
    -0.09
    çīĪ
    -0.09
    æīĢå¾Ĺ
    -0.09
    èĻı
    -0.09
    ueling
    -0.08
    èĭĭ
    -0.08
    USIC
    -0.08
    POSITIVE LOGITS
     tricks
    0.12
     bag
    0.11
    éļIJèĹı
    0.10
    bag
    0.10
    æľ¬äºĭ
    0.10
    è¢ĸ
    0.09
     hidden
    0.09
    éļ±
    0.09
    _hide
    0.09
    è¢ĭ
    0.09
    Activations Density 0.051%

    No Known Activations