© 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. 6731
    Prev
    Next
    INDEX
    Explanations

    move on

    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
     overt
    -0.09
    æĮĤåı·
    -0.09
    å®Įåħ¨æĺ¯
    -0.08
    è´¬
    -0.08
     KH
    -0.08
    çĭİ
    -0.08
    евиÑĩ
    -0.08
    гиб
    -0.08
    èĩ§
    -0.07
    èł¢
    -0.07
    POSITIVE LOGITS
    åīįè¡Į
    0.20
    åīįè¿Ľ
    0.18
    继ç»Ń
    0.17
    ä¸ĭåİ»
    0.17
     continuation
    0.14
    ç¹¼çºĮ
    0.14
     proceed
    0.13
    èµ°ä¸ĭåİ»
    0.13
     continue
    0.13
     ì§Ħíĸī
    0.13
    Activations Density 0.191%

    No Known Activations