© 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. 142231
    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
    eam
    -0.28
    ea
    -0.27
    -stack
    -0.26
    Vm
    -0.25
    ç²¾
    -0.24
    ocrat
    -0.24
    omat
    -0.24
    åħļæĢ»æĶ¯
    -0.24
     Cities
    -0.23
    æ£ķ
    -0.23
    POSITIVE LOGITS
     susp
    0.26
    ä»ĭåħ¥
    0.26
    intval
    0.25
    ivals
    0.25
    ä¸ĵ注
    0.24
    aday
    0.24
     med
    0.24
    åĨ¥
    0.24
    ä¸ĵå±ŀ
    0.24
    å¹²æ¶ī
    0.24
    Activations Density 0.073%

    No Known Activations

    This feature has no known activations.