© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 8315
    Prev
    Next
    INDEX
    Explanations

    preparation and process

    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
    çĤ³
    -0.09
     ble
    -0.09
    ble
    -0.09
    è¯ĦåĪ¤
    -0.08
    æĸ°å¼Ģ
    -0.08
    éĥ¦
    -0.08
    å¿łè¯ļ
    -0.08
    触ãĤĮ
    -0.08
    æį·
    -0.08
     Anc
    -0.08
    POSITIVE LOGITS
    åĩĨå¤ĩ
    0.32
    åĩĨå¤ĩå·¥ä½ľ
    0.29
    åĩĨå¤ĩ好
    0.28
    æºĸåĤĻ
    0.28
     preparation
    0.25
     preparations
    0.23
     prepare
    0.23
    .prepare
    0.22
    _prepare
    0.22
     prepared
    0.22
    Activations Density 0.070%

    No Known Activations