© 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. 30-TRANSCODER-HP
    7. 96207
    Prev
    Next
    INDEX
    Explanations

    Code and data snippets

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    äºij
    -0.32
     cloud
    -0.29
    estyle
    -0.29
    åįķ车
    -0.29
    空ä¸Ń
    -0.27
    äºij端
    -0.27
    éĽ²
    -0.26
    _PO
    -0.26
    ieder
    -0.25
    urn
    -0.25
    POSITIVE LOGITS
     matter
    0.35
    matter
    0.30
     Matter
    0.29
     mattered
    0.28
     listened
    0.26
    'https
    0.26
    _exact
    0.26
    åIJ¬
    0.26
    ť
    0.25
    getDescription
    0.25
    Activations Density 0.881%

    No Known Activations