© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-8B
    5. 18-RESID-BATCHTOPK-65K__L0-80
    7. 3355
    Prev
    Next
    INDEX
    Explanations

    makes it worse

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-8b-saes/saes_Qwen_Qwen3-8B_batch_top_k/resid_post_layer_18
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     Tur
    -0.27
    æĹ¶è¡¨ç¤º
    -0.25
    åħ¼å®¹
    -0.24
     denomination
    -0.23
     Hit
    -0.23
    /left
    -0.22
    åIJĮä»ģ
    -0.22
     Orange
    -0.22
    åľ¨æĹ¥æľ¬
    -0.22
     Dram
    -0.21
    POSITIVE LOGITS
    åıįè¿ĩæĿ¥
    0.55
    è¿Ľä¸ĢæŃ¥
    0.39
    åıįé¦Ī
    0.37
     further
    0.36
    èĩªèº«
    0.35
     itself
    0.34
    æĸ°ä¸Ģè½®
    0.34
    æľ¬èº«
    0.33
     feedback
    0.33
    èĩªèº«çļĦ
    0.32
    Activations Density 0.159%

    No Known Activations