© 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. 52781
    Prev
    Next
    INDEX
    Explanations

    rooms within buildings

    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
    Negative Logits
     directly
    -0.40
    esen
    -0.31
    缴æİ¥
    -0.29
    _direct
    -0.29
    çħĬ
    -0.28
    lius
    -0.28
    é¤į
    -0.26
    Straight
    -0.26
    Direct
    -0.26
     Direct
    -0.26
    POSITIVE LOGITS
    æİ
    0.28
     Cop
    0.28
     surrounds
    0.28
    hrs
    0.28
     salts
    0.27
    ');</
    0.26
    èĪĮå°ĸ
    0.26
    HR
    0.25
    ')</
    0.25
    hr
    0.25
    Activations Density 0.014%

    No Known Activations