© 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. 51078
    Prev
    Next
    INDEX
    Explanations

    drawing attention

    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
    åħįç¨İ
    -0.11
    èµ°å¾Ĺ
    -0.10
    Powered
    -0.10
    çģ«çĤ¬
    -0.10
    騰
    -0.09
    åľ¨åĽ½åĨħ
    -0.09
    èĩĨ
    -0.09
    åݦ
    -0.09
    æīĢå¾Ĺ
    -0.08
    çħ§
    -0.08
    POSITIVE LOGITS
    注æĦıåĬĽ
    0.20
     attention
    0.18
    åIJ¸å¼ķ
    0.16
     Attention
    0.15
    attention
    0.15
    Attention
    0.14
     внимание
    0.13
    åIJ¸å¼ķäºĨ
    0.13
    _attention
    0.12
    åIJ¸å¼ķåĬĽ
    0.12
    Activations Density 0.091%

    No Known Activations