© 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. 22-TRANSCODER-HP
    7. 87483
    Prev
    Next
    INDEX
    Explanations

    code and file formats

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    èĦī
    -0.28
     restitution
    -0.26
    å¦Ŀ
    -0.26
    ona
    -0.25
     lean
    -0.25
     penetration
    -0.24
     parenthesis
    -0.23
     converters
    -0.23
     barren
    -0.23
    競çĪŃ
    -0.23
    POSITIVE LOGITS
    循
    0.25
    semble
    0.25
    游
    0.25
    Unchecked
    0.24
     @@↵
    0.24
    xEC
    0.24
    ottle
    0.23
    æĹłåı¯
    0.23
    sWith
    0.23
     ----------------------------------------------------------------------------↵
    0.23
    Activations Density 0.005%

    No Known Activations