© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 32656
    Prev
    Next
    INDEX
    Explanations

    mutual relationships and concepts

    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
     latter
    -0.09
    liness
    -0.09
    çīĩåŃIJ
    -0.09
    绪
    -0.09
    ÙĴ
    -0.09
    ening
    -0.08
    еÑĤÑĮ
    -0.08
    ableObject
    -0.08
    èĢĮä¸Ķ
    -0.08
    ãģĦãģij
    -0.08
    POSITIVE LOGITS
    ization
    0.14
    ware
    0.12
    ité
    0.11
    istic
    0.11
    ties
    0.10
    åĤį
    0.10
    itarian
    0.10
    æĢ§å¼º
    0.10
    most
    0.10
     зай
    0.10
    Activations Density 0.010%

    No Known Activations