© 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. 50705
    Prev
    Next
    INDEX
    Explanations

    of compressed, flat, addition, turbulence

    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
    åħ¥åı£
    -0.09
    ÂĢÂ
    -0.09
    bug
    -0.09
    æݨèįIJ
    -0.08
    大èĦij
    -0.08
    第äºĮ大
    -0.08
    æķ¬
    -0.08
    Çİ
    -0.08
     Desired
    -0.08
    é²ģ
    -0.08
    POSITIVE LOGITS
    ousse
    0.10
    ä¾Ľåħ»
    0.09
    å¹Ķ
    0.09
    ç¥ŀéĢļ
    0.09
    çĪ±ä¸Ĭ
    0.08
    æĪIJé¾Ļ
    0.08
    åĨħéĻĨ
    0.08
    çĶ©
    0.08
     Sah
    0.08
    è£ħå¤ĩ
    0.08
    Activations Density 0.096%

    No Known Activations