© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 20921
    Prev
    Next
    INDEX
    Explanations

    best way to

    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
    夫
    -0.09
     entertain
    -0.08
     entertained
    -0.08
    CADE
    -0.08
    驾
    -0.08
    çIJ¤
    -0.08
    cul
    -0.08
    åİĨåı²æĢ§
    -0.08
    abl
    -0.08
    Enlarge
    -0.08
    POSITIVE LOGITS
    çļĦåĬŀæ³ķ
    0.15
    æľĢ好çļĦ
    0.14
     way
    0.12
    æľĢéĩįè¦ģ
    0.12
    åĬŀæ³ķ
    0.11
    çļĦæľĢä½³
    0.11
    é¦ĸè¦ģ
    0.11
    åĶ¯ä¸Ģ
    0.11
    æľĢéĩįè¦ģçļĦ
    0.11
    çļĦæł¹æľ¬
    0.10
    Activations Density 0.061%

    No Known Activations