© 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
    2. Qwen3-32B
    3. 32-RESID-BATCHTOPK-65K
    4. 62719
    Prev
    Next
    INDEX
    Explanations

    month. or groundbreaking, award-winning, or the first film to2. methods (usually utilizing thiols or primary amines) are generally expensive, and often require postsynthetic reaction with the DNA3. completion, all graduates (N = 146; 100%) completed an4. ago (like Bridget Jones in her new diary) I couldn’t grasp it at all.methods (usually utilizing thiols or primary amines)

    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
    å®īå¿ĥ
    -0.12
    irie
    -0.10
    çĬĬ
    -0.10
    oven
    -0.10
    çģ
    -0.10
    æŀģåĵģ
    -0.10
    åľ°çIJĥä¸Ĭ
    -0.10
    usal
    -0.10
    _CSR
    -0.09
    è¿Ļæĺ¯
    -0.09
    POSITIVE LOGITS
    为æĤ¨
    0.09
    NT
    0.08
     NT
    0.08
    “
    0.08
    为大家
    0.07
     or
    0.07
    ä¹Łåıªèĥ½
    0.07
    виз
    0.07
    and
    0.07
    )
    0.07
    Activations Density 0.238%

    No Known Activations