© 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. 32173
    Prev
    Next
    INDEX
    Explanations

    Collins and Colin

    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
    /disable
    -0.09
    å®ļäºĨ
    -0.09
    eer
    -0.09
    izon
    -0.08
    饰
    -0.08
    æŀĭ
    -0.08
    åĸ±
    -0.08
    ivalent
    -0.08
    æĶ¾æīĭ
    -0.08
    ç²¾çģµ
    -0.08
    POSITIVE LOGITS
    ses
    0.11
    бо
    0.11
    åĢij
    0.10
    çħ§æł·
    0.10
    utes
    0.10
    worth
    0.10
    nal
    0.10
    ycin
    0.09
    éĹ´éļĶ
    0.09
    ê¹IJ
    0.09
    Activations Density 0.004%

    No Known Activations