© 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. 107905
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    AVA
    -0.28
    idge
    -0.25
    GRAM
    -0.25
    Ï´
    -0.25
    rium
    -0.25
    AFP
    -0.25
    éĩijå±±
    -0.24
    bery
    -0.24
    æĨ¬
    -0.24
    AVOR
    -0.23
    POSITIVE LOGITS
    åİĭ
    0.27
    inic
    0.26
    èĪį
    0.26
    éĹªçĶµ
    0.25
    糸
    0.25
    循çݯ
    0.25
    æľĢå¿«çļĦ
    0.24
     licence
    0.24
     Hughes
    0.24
     Licence
    0.24
    Activations Density 17.437%

    No Known Activations

    This feature has no known activations.