© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 52980
    Prev
    Next
    INDEX
    Explanations

    capture imagination

    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
    ego
    -0.11
    åĽŀè´Ń
    -0.10
    æĶ¶çº³
    -0.09
    æľīèĥ½åĬĽ
    -0.08
    eos
    -0.08
    亲æīĭ
    -0.08
    åį¸
    -0.08
    裸
    -0.08
    æĬ¤
    -0.08
    访
    -0.08
    POSITIVE LOGITS
     susp
    0.10
     дÑĥÑħ
    0.10
    ç²¾é«ĵ
    0.10
    人å¿ĥ
    0.10
     imagination
    0.09
     essence
    0.09
    è¿Ļä¸ĢçĤ¹
    0.09
     attention
    0.08
    åľ°åŁŁ
    0.08
    ç¼ħ
    0.08
    Activations Density 0.014%

    No Known Activations