© 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. 55921
    Prev
    Next
    INDEX
    Explanations

    broken, frozen, chosen

    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
    een
    -0.11
    eenth
    -0.11
    phones
    -0.11
    sWith
    -0.10
    space
    -0.10
    go
    -0.10
     noqa
    -0.10
    mlin
    -0.10
    down
    -0.10
    eners
    -0.10
    POSITIVE LOGITS
    ning
    0.14
    åijIJ
    0.13
    naires
    0.13
    naire
    0.13
    人éĥ½
    0.12
     dolayı
    0.12
    åŃIJéĩĮ
    0.12
    ess
    0.12
    omics
    0.11
    ality
    0.11
    Activations Density 0.037%

    No Known Activations