© 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. 33-TRANSCODER-HP
    7. 52875
    Prev
    Next
    INDEX
    Explanations

    comma

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    irts
    -0.30
    .refs
    -0.26
    .frequency
    -0.23
    iffies
    -0.23
    ypad
    -0.23
    éĢħ
    -0.23
    éĢ¢
    -0.23
     pÅĤyt
    -0.23
    Į¨
    -0.23
    èĭ±å¯¸
    -0.23
    POSITIVE LOGITS
     III
    0.32
     An
    0.31
     Sr
    0.31
     head
    0.29
     II
    0.29
    çļĦæĶ¿æ²»
    0.28
     worked
    0.28
    head
    0.27
     jr
    0.27
     IV
    0.27
    Activations Density 0.108%

    No Known Activations