© 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. 30-TRANSCODER-HP
    7. 97135
    Prev
    Next
    INDEX
    Explanations

    people

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     Prof
    -0.28
    æīĴ
    -0.26
     hangs
    -0.26
    æĹ¢èĥ½
    -0.26
     hanging
    -0.26
     Found
    -0.25
     suspended
    -0.25
    åĬŁèĥ½æĢ§
    -0.25
    currentPage
    -0.25
    tram
    -0.25
    POSITIVE LOGITS
     say
    0.36
    è´¨çĸij
    0.30
     saying
    0.30
    说
    0.30
     debate
    0.29
    say
    0.28
    éĹ®æĪij
    0.28
     ask
    0.28
     estimate
    0.28
     struggle
    0.27
    Activations Density 0.898%

    No Known Activations