© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 142004
    Prev
    Next
    INDEX
    Explanations

    scientific/medical texts

    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
    Not in Any Lists

    No Comments

    Negative Logits
    Dead
    -0.27
     Dead
    -0.26
    »¿
    -0.26
    æŃ»
    -0.25
     DEAD
    -0.24
    å¹³åĿ¦
    -0.23
    genden
    -0.23
    IDER
    -0.23
    sek
    -0.23
    dead
    -0.23
    POSITIVE LOGITS
    fits
    0.29
    ä¹Łè®¸
    0.27
    çĦ¦
    0.27
    æĪĸ许
    0.25
    åı·çłģ
    0.25
    possibly
    0.25
    .py
    0.25
    ross
    0.24
    ä¹Łè®¸æĺ¯
    0.24
    ç®ĹäºĨ
    0.24
    Activations Density 14.559%

    No Known Activations