© 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. 18170
    Prev
    Next
    INDEX
    Explanations

    web code and image tags

    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
    metadata
    -0.28
     IQ
    -0.28
    åħĥ
    -0.28
    ĪæĿĥ
    -0.27
    IQ
    -0.27
     copyrighted
    -0.26
    ä¾µçĬ¯
    -0.26
     metadata
    -0.25
    WER
    -0.25
    _METADATA
    -0.25
    POSITIVE LOGITS
     stagger
    0.27
    -Y
    0.26
    obra
    0.25
    oley
    0.25
    iard
    0.24
    оÐ
    0.24
     yell
    0.24
    edium
    0.23
    ulatory
    0.23
     Txt
    0.23
    Activations Density 0.298%

    No Known Activations