© 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. 27281
    Prev
    Next
    INDEX
    Explanations

    expert followed by descriptor

    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
    å°±è¿Ļä¹Ī
    -0.09
    bourne
    -0.09
    .requireNonNull
    -0.09
    creens
    -0.08
    ä¿¡ä»°
    -0.08
    è§Ĥ念
    -0.08
    åħįçĸ«åĬĽ
    -0.08
    æĻ®æĥł
    -0.08
    裤åŃIJ
    -0.08
    产æĿĥ
    -0.08
    POSITIVE LOGITS
    们
    0.11
    们çļĦ
    0.11
    ly
    0.10
    åĢij
    0.10
    èĪ¬çļĦ
    0.10
     advice
    0.09
    ically
    0.09
    æİ·
    0.09
    ษา
    0.09
     Schultz
    0.09
    Activations Density 0.008%

    No Known Activations