© 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. 51953
    Prev
    Next
    INDEX
    Explanations

    original version comparison

    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
    å§ĭ建
    -0.10
    äºĭåIJİ
    -0.09
    åıĪ好
    -0.09
    3
    -0.08
    æķ´ä½ĵ
    -0.08
    大夫
    -0.08
    æ£Ģ
    -0.08
    åĪĿ级
    -0.08
    æıIJæĹ©
    -0.08
    Wins
    -0.08
    POSITIVE LOGITS
    æĦıä¹īä¸ĬçļĦ
    0.11
    /original
    0.10
    umen
    0.10
    /general
    0.10
    âľ½
    0.10
    æIJł
    0.09
    togroup
    0.09
    åıijæī¬
    0.09
    iterated
    0.09
    spNet
    0.09
    Activations Density 0.027%

    No Known Activations