© 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. 31-TRANSCODER-HP
    7. 18450
    Prev
    Next
    INDEX
    Explanations

    code, math, computing operations

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ard
    -0.27
    åķ·
    -0.26
    athering
    -0.26
    arus
    -0.26
    iere
    -0.26
    eve
    -0.26
    è£ģ
    -0.24
    stand
    -0.24
    ities
    -0.24
     scratched
    -0.24
    POSITIVE LOGITS
    ointed
    0.30
     multiplied
    0.27
    sWith
    0.26
    æ±ģ
    0.26
    妹
    0.26
    ä¹ĺ
    0.25
    åIJ¬åIJ¬
    0.25
    plug
    0.24
    å¤ĩæĪĺ
    0.24
    plied
    0.24
    Activations Density 1.417%

    No Known Activations