© 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. 31-TRANSCODER-HP
    7. 19955
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    å¹¿æ³ĽçļĦ
    -0.27
    athers
    -0.27
    _tokenize
    -0.25
    被æīĵ
    -0.24
    lz
    -0.24
     sophistic
    -0.24
     иÑģÑĤоÑĩник
    -0.24
    Tw
    -0.24
    满满çļĦ
    -0.24
    Having
    -0.23
    POSITIVE LOGITS
    /cgi
    0.28
    åĸĻ
    0.26
    PED
    0.26
    .onSubmit
    0.25
     actionPerformed
    0.25
     dereg
    0.25
    æijĬ
    0.25
    usi
    0.24
    ĻĤ
    0.24
    ä»ħä¾ĽåıĤèĢĥ
    0.24
    Activations Density 0.387%

    No Known Activations