© 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. 19551
    Prev
    Next
    INDEX
    Explanations

    Troy Peterson, Trojan horse, Cottam's Blog

    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
    èĩªå¦Ĥ
    -0.09
    formats
    -0.09
    çĶ¬
    -0.09
     Weinstein
    -0.09
    aru
    -0.09
    orious
    -0.09
    èĩªè´¸åĮº
    -0.09
    æĹ¶è¾°
    -0.09
    ients
    -0.08
    æѤæĸĩ
    -0.08
    POSITIVE LOGITS
    ernet
    0.11
    eur
    0.11
    å§Ķåī¯ä¹¦è®°
    0.10
    Ķ
    0.09
     slick
    0.09
     miá»ĩng
    0.09
    ////////////////////////////////////////////////////////////////
    0.09
    ////////////
    0.09
    é«¡
    0.09
    Ãłn
    0.08
    Activations Density 0.004%

    No Known Activations