© 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. 22392
    Prev
    Next
    INDEX
    Explanations

    without + action/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
     zwar
    -0.09
    aling
    -0.09
    并éĢļè¿ĩ
    -0.09
    好åIJ§
    -0.08
    phy
    -0.08
    aug
    -0.08
     deco
    -0.08
     WV
    -0.08
    bits
    -0.08
    èĻ½çĦ¶
    -0.08
    POSITIVE LOGITS
    çīºçī²
    0.11
    æĹłéĻIJ
    0.10
    è¿ĩäºĨ
    0.10
    ä¸įä¸Ģå®ļ
    0.10
    èī°
    0.09
    LOCKS
    0.09
    ä¹Ĥ
    0.09
    äºį
    0.09
    åłĩ
    0.09
    èĩ³æĸ¼
    0.09
    Activations Density 0.050%

    No Known Activations