© 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. 55487
    Prev
    Next
    INDEX
    Explanations

    accident vs. intention

    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
    oret
    -0.11
    à¸Ħà¸Ńย
    -0.10
    eg
    -0.09
    cord
    -0.09
    riel
    -0.09
    ä»ħ次äºİ
    -0.09
    bie
    -0.08
    loud
    -0.08
    arcy
    -0.08
    çįł
    -0.08
    POSITIVE LOGITS
    ities
    0.10
    æĢ§å¼º
    0.10
    æ·
    0.10
    为ä¹ĭ
    0.09
    æĢ§åĴĮ
    0.09
    åľ°è¯´
    0.09
    ìłģìĿ¸
    0.09
    ITIES
    0.09
    ément
    0.09
    å¾Ģå¤ĸ
    0.09
    Activations Density 0.018%

    No Known Activations