© 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. 18-TRANSCODER-HP
    7. 89598
    Prev
    Next
    INDEX
    Explanations

    community

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    wers
    -0.29
    ãĢij,ãĢIJ
    -0.27
     hydraulic
    -0.25
    positories
    -0.25
    ymb
    -0.24
    å¤ļéĩį
    -0.23
    ymph
    -0.23
     tastes
    -0.23
    è¿Ļåıªæĺ¯
    -0.23
    ender
    -0.23
    POSITIVE LOGITS
    èµ·æĿ¥äºĨ
    0.29
     Corner
    0.26
     poisoning
    0.26
    irl
    0.24
    æĹ¶è¾°
    0.24
     observers
    0.24
     irregular
    0.23
     regulating
    0.23
    æĽ´å¥½åľ°
    0.23
    交代
    0.23
    Activations Density 1.206%

    No Known Activations