© 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. 6507
    Prev
    Next
    INDEX
    Explanations

    constructive alignment, physical, send, exploding, quant, transition

    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
     дÑĥÑħов
    -0.09
    è¸Į
    -0.09
    æĤłéĹ²
    -0.08
    éļ¾
    -0.08
    Nice
    -0.08
    ed
    -0.08
     Nice
    -0.08
    çŀĦåĩĨ
    -0.08
    ãĤĵãģª
    -0.08
    ви
    -0.08
    POSITIVE LOGITS
     practical
    0.13
    å®ŀçĶ¨
    0.13
     пÑĢакÑĤиÑĩеÑģ
    0.12
     Practical
    0.12
    客è§Ĥ
    0.11
    çIJĨæĢ§
    0.11
     concrete
    0.11
    ä¿Ĺ
    0.11
    å®ŀéĻħè¡ĮåĬ¨
    0.11
     technical
    0.10
    Activations Density 0.246%

    No Known Activations