© 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. 64805
    Prev
    Next
    INDEX
    Explanations

    Text snippets and references

    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
    为代表
    -0.26
     NotImplemented
    -0.24
    häuser
    -0.24
    croft
    -0.24
     Closet
    -0.24
    Blocking
    -0.24
    stood
    -0.24
    erah
    -0.23
    обÑī
    -0.23
     sø
    -0.23
    POSITIVE LOGITS
    è¿Ļ对
    0.26
    adera
    0.25
    _UID
    0.25
    çĵ¦
    0.24
    é£İåIJ¹
    0.24
     generations
    0.24
    é£İ
    0.24
    åĽŀé¦Ī
    0.23
    è¿Ļ让
    0.23
    reuse
    0.23
    Activations Density 0.129%

    No Known Activations