© 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. 20-TRANSCODER-HP
    7. 129261
    Prev
    Next
    INDEX
    Explanations

    Allowing/enabling

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    åİ¥
    -0.27
    Saga
    -0.27
    çļĦçłĶåıij
    -0.26
    itty
    -0.25
    Iterations
    -0.25
    npos
    -0.25
    çĿ£
    -0.24
    æµģåĬ¨
    -0.24
     Georgian
    -0.24
    -Identifier
    -0.23
    POSITIVE LOGITS
    åľ¨ä»ĸçļĦ
    0.30
     lenses
    0.25
    elib
    0.24
     belts
    0.24
     Lens
    0.24
    midd
    0.24
    rors
    0.24
    enze
    0.24
    men
    0.23
     diam
    0.23
    Activations Density 27.188%

    No Known Activations