© 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. 47839
    Prev
    Next
    INDEX
    Explanations

    company program options details

    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
    ãī
    -0.08
    лож
    -0.08
    ful
    -0.08
    åİŁåĽłæĺ¯
    -0.08
    less
    -0.08
    éĹ®é¢ĺæĺ¯
    -0.08
    poi
    -0.08
    ense
    -0.08
    ableObject
    -0.08
    çĬ¶æĢģä¸ĭ
    -0.08
    POSITIVE LOGITS
    产ä¸ļåĮĸ
    0.10
    庶
    0.09
    AllWindows
    0.09
    å®Ĺ
    0.09
    çĤİ
    0.09
    åı¦ä¸Ģä½į
    0.08
    itere
    0.08
    ãģªãĤĭ
    0.08
    åĴĮ个人
    0.08
    å¤§åľ°
    0.08
    Activations Density 4.849%

    No Known Activations