© 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. 33-TRANSCODER-HP
    7. 53854
    Prev
    Next
    INDEX
    Explanations

    say "more"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     amount
    -0.32
     Amount
    -0.29
    amount
    -0.25
    Amount
    -0.25
     amounts
    -0.25
     concentrating
    -0.24
     drunken
    -0.23
     cent
    -0.23
    pton
    -0.23
    OfWork
    -0.23
    POSITIVE LOGITS
     more
    0.42
    æĽ´åħ·
    0.39
     far
    0.39
    æĽ´å¤ļ
    0.37
    more
    0.35
     better
    0.35
    æĽ´å¥½
    0.34
     further
    0.34
    æĽ´å¤ļçļĦ
    0.34
    -m
    0.34
    Activations Density 0.165%

    No Known Activations