© 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. 17943
    Prev
    Next
    INDEX
    Explanations

    ghost, image, memory, echo

    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
    uncan
    -0.10
    常æĢģ
    -0.09
     Robin
    -0.09
    çªĺ
    -0.09
    vinc
    -0.09
    ills
    -0.08
    _CHARSET
    -0.08
    æĩĴå¾Ĺ
    -0.08
    家åĬ¡
    -0.08
     Rule
    -0.08
    POSITIVE LOGITS
    å½±
    0.12
    身影
    0.12
    çļĦ身影
    0.11
    形象
    0.11
    zial
    0.10
    çļĦ形象
    0.10
    踪
    0.10
     ghost
    0.10
    å½±åĥı
    0.10
    诱
    0.09
    Activations Density 0.087%

    No Known Activations