© 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. 18654
    Prev
    Next
    INDEX
    Explanations

    say "grade"

    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
    med
    -0.28
    inta
    -0.28
    QA
    -0.27
    modo
    -0.27
    ¥
    -0.27
    .layers
    -0.25
     syn
    -0.24
     teachers
    -0.23
     pass
    -0.23
    loh
    -0.23
    POSITIVE LOGITS
    æĪIJåĵģ
    0.28
    BEL
    0.26
    æľĢ好æĺ¯
    0.26
     odd
    0.24
    sville
    0.24
    -log
    0.24
    partment
    0.24
     academic
    0.24
    åĴĮæĶ¯æĮģ
    0.23
    çıł
    0.23
    Activations Density 0.039%

    No Known Activations