© 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. 106993
    Prev
    Next
    INDEX
    Explanations

    choice

    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
    ariate
    -0.27
     (("
    -0.26
     zab
    -0.25
    ortic
    -0.25
    htag
    -0.24
    tit
    -0.24
    æīĢæľī人
    -0.24
    ikit
    -0.23
     [["
    -0.23
    å²ģ以ä¸Ĭ
    -0.23
    POSITIVE LOGITS
     choice
    0.69
    choice
    0.58
     Choice
    0.57
    -choice
    0.56
    _choice
    0.55
    Choice
    0.53
     choosing
    0.52
    éĢīæĭ©
    0.51
    (choice
    0.49
     choose
    0.47
    Activations Density 0.898%

    No Known Activations