© 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. 10252
    Prev
    Next
    INDEX
    Explanations

    questions involving how or what

    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
    碰åĪ°
    -0.09
    paren
    -0.09
    érique
    -0.08
    arent
    -0.08
    èī¯
    -0.08
    åľ¨åIJĮä¸Ģ
    -0.08
    iglia
    -0.08
    ড
    -0.08
    ogh
    -0.08
    aphore
    -0.07
    POSITIVE LOGITS
    å¦Ĥä½ķ
    0.20
     how
    0.20
    å¦Ĥä½ķåİ»
    0.17
     какие
    0.16
    æĢİæł·
    0.16
    ä½ķæŶ
    0.15
    å¦Ĥä½ķçľĭå¾ħ
    0.15
    æĢİä¹Īåģļ
    0.14
    ä»Ģä¹Īæĺ¯
    0.14
     whether
    0.14
    Activations Density 0.121%

    No Known Activations