© 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. 18-TRANSCODER-HP
    7. 89121
    Prev
    Next
    INDEX
    Explanations

    Punctuation and common words

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     unavoidable
    -0.34
    iffe
    -0.27
    å¥ĺ
    -0.26
    iff
    -0.26
    æĹłåı¯
    -0.26
    eman
    -0.26
    éĢł
    -0.26
    åĽ°
    -0.26
     acceptable
    -0.25
     apology
    -0.25
    POSITIVE LOGITS
    vore
    0.25
    移交
    0.25
    bit
    0.25
    çľĭè¿ĩ
    0.24
     onTouch
    0.24
    æĢ»å±Ģ
    0.24
    .pag
    0.24
     ../
    0.23
     enriched
    0.23
    姿
    0.23
    Activations Density 31.536%

    No Known Activations