© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 19830
    Prev
    Next
    INDEX
    Explanations

    Gl or gl

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    祷
    -0.28
    ãĥģãĥ³
    -0.27
    ÑģÑĤи
    -0.26
    主ä¹ī
    -0.25
    /apt
    -0.24
    å°¼
    -0.24
    åĽĽåŃ£
    -0.24
    packages
    -0.24
    INK
    -0.23
    package
    -0.23
    POSITIVE LOGITS
     %+
    0.26
    illon
    0.25
    onda
    0.25
     Pike
    0.25
    elly
    0.25
    oti
    0.25
    éĶIJ
    0.25
    ota
    0.25
    Eb
    0.25
    éĴī
    0.25
    Activations Density 0.959%

    No Known Activations