© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 52416
    Prev
    Next
    INDEX
    Explanations

    assume that

    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
    ening
    -0.11
    usz
    -0.09
    ened
    -0.09
    anco
    -0.08
    ues
    -0.08
    edata
    -0.08
    liness
    -0.08
    buch
    -0.08
    çĺ¾
    -0.08
    ÑĦоÑĢ
    -0.08
    POSITIVE LOGITS
    ably
    0.14
    大家éĥ½åľ¨
    0.12
    ioni
    0.11
    /im
    0.10
    overs
    0.10
    人人éĥ½
    0.10
    è¿ĩéĶĻ
    0.09
    ãģ¹ãģį
    0.09
    责任
    0.09
    Ø«
    0.09
    Activations Density 0.022%

    No Known Activations