© 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. 25-TRANSCODER-HP
    7. 109567
    Prev
    Next
    INDEX
    Explanations

    hot

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    _boolean
    -0.28
     Weiss
    -0.27
     Hess
    -0.27
    ä½İäºİ
    -0.25
    BOOLE
    -0.25
    é¦ĸ家
    -0.25
    LETTE
    -0.25
     paraph
    -0.25
     Boolean
    -0.24
     BOOLEAN
    -0.24
    POSITIVE LOGITS
    hot
    0.28
    çĤ¹çĩĥ
    0.27
    powered
    0.27
    çķĮçļĦ
    0.26
    æĭ¦
    0.26
    æī¹
    0.26
    缼
    0.25
    åĪ°æĿ¥
    0.25
    åŃ£
    0.25
    вали
    0.25
    Activations Density 0.908%

    No Known Activations