© 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. 54376
    Prev
    Next
    INDEX
    Explanations

    replace or replacement

    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
    çŀĦ
    -0.10
    çĭİ
    -0.08
     мн
    -0.08
    .mock
    -0.08
    Lots
    -0.08
    bery
    -0.08
    à¹Ģà¸ļ
    -0.08
    è¿ĺä¸įå¦Ĥ
    -0.08
    aw
    -0.08
     bem
    -0.08
    POSITIVE LOGITS
    æĽ¿æį¢
    0.17
     replacements
    0.16
     replacement
    0.16
    æĽ´æį¢
    0.15
     replace
    0.14
     Replacement
    0.14
    _REPLACE
    0.14
    Replace
    0.13
     Replace
    0.13
    replace
    0.12
    Activations Density 0.099%

    No Known Activations