© 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. 35-TRANSCODER-HP
    7. 39774
    Prev
    Next
    INDEX
    Explanations

    Technical jargon

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åľŁ
    -0.30
    æ¦Ĥçİĩ
    -0.29
    模
    -0.29
    issa
    -0.29
    probability
    -0.28
    çĶµç£ģ
    -0.28
    æīĭåĨĮ
    -0.28
    training
    -0.27
    æıIJåįĩ
    -0.27
     Targets
    -0.27
    POSITIVE LOGITS
     vast
    0.39
     largely
    0.34
    æľīä»·åĢ¼
    0.33
     otherwise
    0.31
     radically
    0.30
    æĥĬ人
    0.30
    çĺł
    0.30
     formidable
    0.30
     predominantly
    0.30
     dwind
    0.30
    Activations Density 68.454%

    No Known Activations