© 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. 62600
    Prev
    Next
    INDEX
    Explanations

    upper portion 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
    ommen
    -0.10
    ead
    -0.09
    èľ£
    -0.09
    ç¿ħ
    -0.09
    omi
    -0.09
    åĨħå¤ĸ
    -0.09
    iec
    -0.09
    èŁª
    -0.08
    çĢļ
    -0.08
    side
    -0.08
    POSITIVE LOGITS
    -upper
    0.10
     ech
    0.09
    íĬ¼
    0.09
    ä¸īå±Ĥ
    0.08
    /right
    0.08
    icker
    0.08
    -desktop
    0.08
    PCS
    0.08
    ıyla
    0.08
    -most
    0.08
    Activations Density 0.102%

    No Known Activations