© 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. 10820
    Prev
    Next
    INDEX
    Explanations

    bowl of

    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
    oral
    -0.11
    ëį°
    -0.09
    opp
    -0.08
     overse
    -0.08
    sst
    -0.08
    رت
    -0.08
     convict
    -0.08
    çĨĬçĮ«
    -0.08
    iros
    -0.08
    oha
    -0.08
    POSITIVE LOGITS
    çŃ·
    0.13
    ings
    0.13
    ongan
    0.11
    iers
    0.10
    ×Ļק
    0.10
    iness
    0.10
    ter
    0.10
    ä»·
    0.09
    .Unsupported
    0.09
    ware
    0.09
    Activations Density 0.022%

    No Known Activations