© 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. 22745
    Prev
    Next
    INDEX
    Explanations

    take advantage of

    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
    é³ĸ
    -0.09
    æĪIJé¾Ļ
    -0.09
    æ¼±
    -0.09
    isz
    -0.09
     Baxter
    -0.09
    æĭĴ
    -0.08
    éģ¥è¿ľ
    -0.08
    бой
    -0.08
    åijĽ
    -0.08
    tml
    -0.08
    POSITIVE LOGITS
    åħħåĪĨåĪ©çĶ¨
    0.22
     aprove
    0.22
    åĪ©çĶ¨
    0.19
     táºŃn
    0.15
     harness
    0.15
     advantage
    0.14
     Harness
    0.14
     exploited
    0.13
     leverage
    0.13
    Harness
    0.13
    Activations Density 0.084%

    No Known Activations