© 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. 20-TRANSCODER-HP
    7. 108129
    Prev
    Next
    INDEX
    Explanations

    technical writing

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĪĹ
    -0.30
    uç
    -0.27
    举
    -0.27
    uran
    -0.27
    urd
    -0.27
    urga
    -0.26
    arged
    -0.26
    player
    -0.26
     Sphinx
    -0.25
    éļ¾
    -0.24
    POSITIVE LOGITS
     Trib
    0.26
    éĴ±
    0.25
    è¾ħåĬ©
    0.24
    _SY
    0.24
    .xr
    0.24
    ç»
    0.23
     awaken
    0.23
    went
    0.23
    éĮ¢
    0.23
    çĽ¸äº²
    0.23
    Activations Density 25.937%

    No Known Activations