© 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. 54648
    Prev
    Next
    INDEX
    Explanations

    Jake Owen

    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
    amp
    -0.17
    ties
    -0.10
    owns
    -0.09
    çĿ¹
    -0.09
    éĩįè¦ģæĦıä¹ī
    -0.09
    cope
    -0.09
    ç¼Ł
    -0.09
    à¹īà¸Ńà¸Ļ
    -0.09
    iron
    -0.08
    æŀ¸
    -0.08
    POSITIVE LOGITS
    sWith
    0.13
    zelf
    0.12
    ement
    0.11
    swith
    0.11
    emente
    0.11
    пи
    0.10
    etwork
    0.09
    al
    0.09
    croft
    0.09
    aled
    0.09
    Activations Density 0.169%

    No Known Activations