© 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. 54706
    Prev
    Next
    INDEX
    Explanations

    sync operations

    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
    æªIJ
    -0.10
    ting
    -0.09
    cmp
    -0.09
    羣åĪĩ
    -0.08
    issa
    -0.08
    urer
    -0.08
    InstanceOf
    -0.08
    tempt
    -0.08
    欢åĸľ
    -0.08
    auce
    -0.08
    POSITIVE LOGITS
    ron
    0.13
    icut
    0.12
    rh
    0.12
    ingly
    0.10
    ableObject
    0.10
    åķľ
    0.10
    fusion
    0.09
    enate
    0.09
    HING
    0.09
    ra
    0.09
    Activations Density 0.051%

    No Known Activations