© 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. 35-TRANSCODER-HP
    7. 39914
    Prev
    Next
    INDEX
    Explanations

    say "e" words

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĥı
    -0.29
    çľĭä¸Ģä¸ĭ
    -0.29
    ä¸įæĸŃåľ°
    -0.28
     iso
    -0.25
    åĽłä¸ºå®ĥ
    -0.24
     highlight
    -0.24
     weak
    -0.24
     forced
    -0.23
    ŀ
    -0.23
     "*",
    -0.23
    POSITIVE LOGITS
    p
    0.32
    paced
    0.30
    piler
    0.29
    pies
    0.29
    encies
    0.28
    pel
    0.28
    arily
    0.28
    ist
    0.28
     padd
    0.27
    pag
    0.27
    Activations Density 4.726%

    No Known Activations