© 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. 26761
    Prev
    Next
    INDEX
    Explanations

    ground sinking or collapsing

    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
    _barrier
    -0.11
     franca
    -0.11
    æĸ§
    -0.10
    odal
    -0.10
    é£Ļ
    -0.09
    .Preference
    -0.09
    askell
    -0.09
    ESC
    -0.09
    arded
    -0.09
    åįĬ天
    -0.09
    POSITIVE LOGITS
    å¡Į
    0.14
     collapsed
    0.12
     collapse
    0.11
    åĿį
    0.11
     collapsing
    0.10
    è¸ı
    0.10
    Collapse
    0.10
    è¸Ł
    0.09
    è¹ĭ
    0.09
     Collapse
    0.09
    Activations Density 0.028%

    No Known Activations