© 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. 1540
    Prev
    Next
    INDEX
    Explanations

    take / capture

    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
    çŃīåĢĻ
    -0.09
     PD
    -0.09
    让å®ĥ
    -0.08
    为æĤ¨
    -0.08
    åľ¨è¿ĩ渡
    -0.08
    èµĦ
    -0.08
    æĽ²æĬĺ
    -0.08
    çıĪ
    -0.08
     Wikimedia
    -0.08
    ells
    -0.08
    POSITIVE LOGITS
    丢失
    0.13
    éĺµåľ°
    0.11
    ä¸Ģ座
    0.11
    é³Į
    0.10
    驾çħ§
    0.10
    夺å¾Ĺ
    0.10
    夺
    0.09
     Capture
    0.09
    çªĿ
    0.09
    å¸Ĥåľºä»½é¢Ŀ
    0.09
    Activations Density 0.103%

    No Known Activations