© 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. 57658
    Prev
    Next
    INDEX
    Explanations

    destroy / destruction

    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
    ients
    -0.09
    ainment
    -0.09
    æĴ·
    -0.09
    è¡Įäºĭ
    -0.09
    enced
    -0.09
    河水
    -0.08
    .collection
    -0.08
    æ¯ĵ
    -0.08
    _exclude
    -0.08
    æ·Ļ
    -0.08
    POSITIVE LOGITS
    æij§æ¯ģ
    0.12
    æ¯ģ
    0.12
    åİŁæľīçļĦ
    0.12
     Destroy
    0.11
    Destroy
    0.11
    .destroy
    0.11
    çļĦä¿¡å¿ĥ
    0.11
     destroy
    0.11
    åİŁæľī
    0.10
    å¨ĩ
    0.10
    Activations Density 0.099%

    No Known Activations