© 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. 32050
    Prev
    Next
    INDEX
    Explanations

    will or would followed by a verb

    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
    æĺ¯çľŁçļĦ
    -0.10
    éĽĴ
    -0.09
     anc
    -0.09
    ä¼ļ被
    -0.09
    unks
    -0.09
    éħĨ
    -0.08
    å§ĭç»ĪåĿļæĮģ
    -0.08
    æ¤į
    -0.08
    éĤł
    -0.08
    ldr
    -0.08
    POSITIVE LOGITS
    å²Ĺ
    0.09
    åŁ¹
    0.08
    å©
    0.08
    æłĩ
    0.08
    è´Ŀå°Ķ
    0.08
    .Pixel
    0.08
    ugged
    0.08
    éĩĩåıĸ
    0.08
    æ±
    0.08
    éĴĵé±¼
    0.08
    Activations Density 0.233%

    No Known Activations