© 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. 49238
    Prev
    Next
    INDEX
    Explanations

    the goal, aim, or intention

    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
    RESS
    -0.10
    etten
    -0.10
    esco
    -0.09
    æŶ髦
    -0.09
    æŶ任
    -0.09
    еÑĢ
    -0.09
    epar
    -0.09
    æŀī
    -0.08
    riad
    -0.08
    ä»Ģä¹Īåľ°æĸ¹
    -0.08
    POSITIVE LOGITS
     regard
    0.20
    standing
    0.19
     regards
    0.18
    stood
    0.15
    outh
    0.13
    drawable
    0.12
     respect
    0.12
    ering
    0.12
     hopes
    0.11
    holding
    0.10
    Activations Density 0.078%

    No Known Activations