© 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. 35912
    Prev
    Next
    INDEX
    Explanations

    negation followed by an action

    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
    åIJ¹
    -0.09
    è¦ģçľĭ
    -0.09
    åĢį
    -0.09
    ecz
    -0.09
    bjerg
    -0.09
    å¿Ĺ强
    -0.08
    年第
    -0.08
    ä¿Ł
    -0.08
    èĻļå¼±
    -0.08
    çªģ
    -0.08
    POSITIVE LOGITS
    å°±ä¸į
    0.13
    å°±ä¸įä¼ļ
    0.11
    就没
    0.10
    就没æľī
    0.10
    æĶ¾æīĭ
    0.09
    Checksum
    0.09
    åĭī强
    0.09
    置身
    0.09
    parm
    0.08
     вообÑīе
    0.08
    Activations Density 0.068%

    No Known Activations