© 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. 47942
    Prev
    Next
    INDEX
    Explanations

    data reduction, data fields, data communication

    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
    车身
    -0.09
    avel
    -0.09
    大å®Ĺ
    -0.09
    ä¸Ģ件äºĭ
    -0.08
    pal
    -0.08
    brand
    -0.08
    ä¿¡æģ¯åıijå¸ĥ
    -0.08
    丼
    -0.08
    åľºåľ°
    -0.08
    odia
    -0.08
    POSITIVE LOGITS
    ë§Ŀ
    0.11
    (Data
    0.10
    éĢıè§Ĩ
    0.10
    /Data
    0.10
    侥幸
    0.10
     pdata
    0.10
    iled
    0.09
    fono
    0.09
    ene
    0.09
     setData
    0.09
    Activations Density 0.297%

    No Known Activations