© 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. 107754
    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
    avor
    -0.28
    åıijå±ķ模å¼ı
    -0.26
     Defensive
    -0.25
     Weight
    -0.24
    许
    -0.24
    -tools
    -0.24
    .cv
    -0.24
     Apex
    -0.24
     queen
    -0.23
    latable
    -0.23
    POSITIVE LOGITS
    粪
    0.28
    ä½Ļ
    0.27
    ç³Ł
    0.25
    ä½İ
    0.24
    èµĦ产
    0.24
    ((&
    0.24
    çŁ¿çī©
    0.24
     {:?}",
    0.24
    å¤ļä½Ļ
    0.24
    踵
    0.24
    Activations Density 89.817%

    No Known Activations

    This feature has no known activations.