© 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. 63022
    Prev
    Next
    INDEX
    Explanations

    twisted, distorted, perverseness, crooked

    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
    ordo
    -0.10
    ooks
    -0.09
     Morrow
    -0.09
    _hz
    -0.08
    uci
    -0.08
    atically
    -0.08
     bordered
    -0.08
    anges
    -0.08
    -in
    -0.08
     Luna
    -0.08
    POSITIVE LOGITS
    à¸Ľà¸ģ
    0.12
    èĻ¬
    0.10
    ä¸įæ¸ħ
    0.10
    PointerException
    0.10
    åĦ¿
    0.09
    ities
    0.09
    鬲
    0.09
    æŃ£å¸¸çļĦ
    0.09
    åħīçħ§
    0.09
    常æĢģ
    0.08
    Activations Density 0.027%

    No Known Activations