© 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. 19-TRANSCODER-HP
    7. 8957
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åIJĦç±»
    -0.28
    emento
    -0.26
     clos
    -0.26
     sug
    -0.25
     recall
    -0.25
    terior
    -0.24
     reap
    -0.24
    èĭķ
    -0.23
    鬼
    -0.23
    ceive
    -0.23
    POSITIVE LOGITS
    éĬ®
    0.26
     мл
    0.26
    XR
    0.25
    foot
    0.25
    adem
    0.25
     Riy
    0.25
    çIJ¬
    0.25
    olders
    0.24
    尤为
    0.24
     Scrolls
    0.24
    Activations Density 0.004%

    No Known Activations

    This feature has no known activations.