© 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. 22-TRANSCODER-HP
    7. 88259
    Prev
    Next
    INDEX
    Explanations

    Champagne

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    would
    -0.59
     would
    -0.56
    Would
    -0.51
     Would
    -0.50
    could
    -0.44
    å°Ĩä¼ļ
    -0.44
     würde
    -0.42
    å°±ä¼ļ
    -0.42
     wouldn
    -0.41
     could
    -0.41
    POSITIVE LOGITS
    Refer
    0.28
    追ãģĦ
    0.26
    交代
    0.25
    ordinate
    0.25
    \.
    0.24
     orientations
    0.24
    ocused
    0.24
     Refer
    0.23
     Acer
    0.23
    ante
    0.23
    Activations Density 0.004%

    No Known Activations