© 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. 12-TRANSCODER-HP
    7. 22044
    Prev
    Next
    INDEX
    Explanations

    roman numerals and '4'

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_12.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.12.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    å§Ķç»Ħç»ĩ
    -0.26
    çļĦåIJĦç§į
    -0.26
    agram
    -0.25
     Views
    -0.25
    å±ī
    -0.24
     sop
    -0.23
    eut
    -0.23
     setObject
    -0.23
    )view
    -0.23
    \s
    -0.23
    POSITIVE LOGITS
    说ä»Ģä¹Ī
    0.28
    림
    0.27
    reece
    0.27
    rait
    0.26
    å¿ĺè®°
    0.25
    èĵij
    0.24
     capitals
    0.24
    elage
    0.23
    iland
    0.23
    ume
    0.23
    Activations Density 0.011%

    No Known Activations