© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 22003
    Prev
    Next
    INDEX
    Explanations

    smell of

    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
    ../../../
    -0.12
    eer
    -0.11
    каÑĤ
    -0.10
    esco
    -0.09
    umph
    -0.09
    ettle
    -0.09
    rovers
    -0.09
    eur
    -0.09
    uggy
    -0.08
    ìĦĿ
    -0.08
    POSITIVE LOGITS
    ander
    0.10
    éĿ¥
    0.10
    dek
    0.10
    æĻ°
    0.10
    paces
    0.10
    te
    0.10
    踪
    0.09
    ×ķת
    0.09
    åĭĥåĭĥ
    0.09
    time
    0.09
    Activations Density 0.005%

    No Known Activations