© 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. 23248
    Prev
    Next
    INDEX
    Explanations

    suite

    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
     parte
    -0.10
    éĵ®
    -0.09
    r
    -0.08
    èĮ¬
    -0.08
    sWith
    -0.08
     окÑĢÑĥжа
    -0.08
    thic
    -0.08
    sign
    -0.08
    wel
    -0.08
    ç²īå°ĺ
    -0.08
    POSITIVE LOGITS
    åĨħçļĦ
    0.10
    anca
    0.09
    ehler
    0.09
    orce
    0.09
    aders
    0.09
    ties
    0.09
    elor
    0.09
    enge
    0.09
    agram
    0.09
    eners
    0.09
    Activations Density 0.007%

    No Known Activations