© 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. 34675
    Prev
    Next
    INDEX
    Explanations

    diagnostic accuracy and metrics

    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
    许
    -0.10
    èĺ¸
    -0.10
    ä»ĺ
    -0.10
    å¹²
    -0.09
     rhet
    -0.09
    bler
    -0.09
    驯
    -0.09
    缴åįĩ
    -0.09
    ITE
    -0.08
    å¸ĮæľĽéĢļè¿ĩ
    -0.08
    POSITIVE LOGITS
    åĩĨç¡®æĢ§
    0.11
     singly
    0.10
     accuracy
    0.10
    accuracy
    0.09
    ovo
    0.09
     ferr
    0.09
    ION
    0.09
     Accuracy
    0.09
    è¿Ľåĩº
    0.08
    ctp
    0.08
    Activations Density 0.377%

    No Known Activations