© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 22-TRANSCODER-HP
    7. 110993
    Prev
    Next
    INDEX
    Explanations

    Code snippets

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ocrat
    -0.28
    çĶŁäº§åĴĮ
    -0.27
    dera
    -0.27
     NTN
    -0.27
    oni
    -0.26
     //<
    -0.26
     sucker
    -0.25
    ruz
    -0.25
    .multi
    -0.25
    æľīå¤ļ
    -0.25
    POSITIVE LOGITS
    åĥµ
    0.30
    æľªèĥ½
    0.29
    jaw
    0.29
     unintended
    0.27
    æ¯Ľ
    0.27
     silent
    0.26
    ä¸įäºĨ
    0.26
    algo
    0.25
     programs
    0.25
     wrong
    0.25
    Activations Density 0.002%

    No Known Activations