© 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. 19975
    Prev
    Next
    INDEX
    Explanations

    and

    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
    (handles
    -0.30
     Inserts
    -0.25
    ä¸Ģå°ģ
    -0.25
    /apps
    -0.25
    ä¸įå¿į
    -0.24
    èĪ³
    -0.24
     paid
    -0.24
    åijĪ
    -0.24
     Paid
    -0.23
    èĩªç§°
    -0.23
    POSITIVE LOGITS
     monitor
    0.42
    è·Łè¸ª
    0.41
    çĽijæµĭ
    0.39
    追踪
    0.38
     assess
    0.38
     analyze
    0.38
     tracking
    0.37
    éī´å®ļ
    0.36
     evaluate
    0.36
     track
    0.35
    Activations Density 3.114%

    No Known Activations