© 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. 22063
    Prev
    Next
    INDEX
    Explanations

    hurt, offended, offence

    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
    åĤ¬
    -0.09
    iman
    -0.09
    fall
    -0.08
    宣
    -0.08
    каÑı
    -0.08
     dab
    -0.08
    SPAN
    -0.08
    å¹²äºĭ
    -0.08
     rever
    -0.08
    ä¸įåģľ
    -0.08
    POSITIVE LOGITS
    æĦŁåĬ¨
    0.11
     Braun
    0.10
    Convention
    0.10
    æĦŁæ¿Ģ
    0.09
    egra
    0.09
    对æĸ¹
    0.09
    èIJľ
    0.09
    esan
    0.09
    黩
    0.09
    umu
    0.09
    Activations Density 0.063%

    No Known Activations