© 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. 32163
    Prev
    Next
    INDEX
    Explanations

    "well" / "good" indicating state

    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
    è¾Ľ
    -0.11
    quer
    -0.10
    ahn
    -0.10
    åłĥ
    -0.09
    踽
    -0.08
    bens
    -0.08
    认è´Ń
    -0.08
    ä¹Łä¸įè¦ģ
    -0.08
     jung
    -0.08
    åģľå·¥
    -0.08
    POSITIVE LOGITS
    å¾Ĺ好
    0.19
    å¾Ĺå¾Ī好
    0.19
    èī¯å¥½
    0.18
    èī¯å¥½çļĦ
    0.16
    å¾Ī好
    0.15
    æĮºå¥½
    0.15
    è¿ĺä¸įéĶĻ
    0.14
    好çļĦ
    0.14
    è¶Ĭ好
    0.14
    éĿŀ常好
    0.14
    Activations Density 0.069%

    No Known Activations