© 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. 32-TRANSCODER-HP
    7. 41693
    Prev
    Next
    INDEX
    Explanations

    source code/licenses

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åħ¶ä»ĸ
    -0.27
     other
    -0.27
    LETTE
    -0.26
     Pru
    -0.26
    å¸ĤåĢ¼
    -0.26
    ecn
    -0.26
    _metrics
    -0.25
    cret
    -0.25
    社ä¼ļç§ijåѦ
    -0.24
    coverage
    -0.24
    POSITIVE LOGITS
    çIJĨäºĭ
    0.30
    ãģ¶
    0.26
    rum
    0.26
    èµ°å¾Ĺ
    0.26
    hoot
    0.25
    ÑĦÑĦ
    0.25
     reserv
    0.25
    è¶Ĭé«ĺ
    0.24
     flick
    0.24
    æĶ
    0.24
    Activations Density 0.003%

    No Known Activations