© 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. 17686
    Prev
    Next
    INDEX
    Explanations

    paramount importance or extremely important

    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
    ä¸įåħ·å¤ĩ
    -0.09
    屯
    -0.09
    ä¸įåIJ«
    -0.09
    oria
    -0.08
    çıī
    -0.08
    ResponseType
    -0.08
    åīĤéĩı
    -0.08
    éªĭ
    -0.08
    èİĴ
    -0.08
    èijĽ
    -0.08
    POSITIVE LOGITS
    éĿŀ常éĩįè¦ģ
    0.28
    å¾Īéĩįè¦ģ
    0.27
     importance
    0.27
    éĩįè¦ģ
    0.27
     важ
    0.26
    åįģåĪĨéĩįè¦ģ
    0.26
    èĩ³åħ³éĩįè¦ģ
    0.26
    çļĦéĩįè¦ģæĢ§
    0.26
     important
    0.26
    éĩįè¦ģçļĦ
    0.24
    Activations Density 0.175%

    No Known Activations