© 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. 30-TRANSCODER-HP
    7. 96872
    Prev
    Next
    INDEX
    Explanations

    generate

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     generating
    -1.05
     generate
    -1.05
    çĶŁæĪIJ
    -1.05
     generates
    -1.02
    产çĶŁ
    -0.98
    产çĶŁçļĦ
    -0.95
     generated
    -0.93
     Generate
    -0.93
    generate
    -0.90
    Generate
    -0.90
    POSITIVE LOGITS
    porte
    0.28
     porta
    0.26
    porter
    0.26
    odge
    0.25
    éļ¾
    0.24
    vale
    0.24
     ports
    0.24
    æł·æľ¬
    0.23
    ogram
    0.23
    æł·
    0.23
    Activations Density 0.907%

    No Known Activations