© 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. 17413
    Prev
    Next
    INDEX
    Explanations

    bearded man

    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
    ми
    -0.10
     اÙĦØŃر
    -0.09
    è¯ĥ
    -0.09
    ç¹ģ
    -0.09
    ä¸įäºĨ
    -0.09
     unnamed
    -0.09
    驾é©Ń
    -0.09
    krit
    -0.09
     Arbor
    -0.09
    aneously
    -0.08
    POSITIVE LOGITS
    edly
    0.11
    acle
    0.10
    lessness
    0.10
    ä¸į解
    0.10
    ings
    0.09
    iness
    0.09
    agers
    0.09
    LESS
    0.09
    èī®
    0.09
    icit
    0.08
    Activations Density 0.004%

    No Known Activations