© 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. 12-TRANSCODER-HP
    7. 23295
    Prev
    Next
    INDEX
    Explanations

    generated

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_12.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.12.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     orient
    -0.29
    éĺħè§Ī
    -0.27
     therein
    -0.27
     recalling
    -0.26
     recall
    -0.26
    VISION
    -0.26
    å¯ĴåĨ¬
    -0.26
    icum
    -0.25
    illion
    -0.24
     cen
    -0.24
    POSITIVE LOGITS
    转åĮĸ
    0.29
    ç¿»è¯ij
    0.28
    arkin
    0.27
     Coverage
    0.27
    æµģ转
    0.26
    转åĮĸ为
    0.26
    ë¹Į
    0.25
    αι
    0.25
    åı¯æĥľ
    0.25
     translated
    0.24
    Activations Density 0.007%

    No Known Activations