© 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. 18775
    Prev
    Next
    INDEX
    Explanations

    mystery, mysterious, box, film, secret, Benedict

    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
    manship
    -0.12
    yard
    -0.09
    itional
    -0.09
    åķĸ
    -0.09
    rock
    -0.08
    icularly
    -0.08
    åĢĮ
    -0.08
    ç¨İåĬ¡å±Ģ
    -0.08
    ycle
    -0.08
    -off
    -0.08
    POSITIVE LOGITS
    è§ĴèIJ½
    0.11
    åĭĥåĭĥ
    0.11
    ellaneous
    0.10
    æľªçŁ¥
    0.10
    ariat
    0.10
    çħ§æĸĻ
    0.10
    åħ®
    0.09
    èĺŃ
    0.09
    isk
    0.09
    åIJ½
    0.09
    Activations Density 0.009%

    No Known Activations