© 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. 19730
    Prev
    Next
    INDEX
    Explanations

    Code

    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
    åİī
    -0.30
     posted
    -0.29
    éĻIJæŶ
    -0.27
    _levels
    -0.26
    说ä»Ģä¹Ī
    -0.25
    çŁŃ线
    -0.25
    ulary
    -0.24
    ä¸ķ
    -0.24
    åłī
    -0.24
    -bound
    -0.23
    POSITIVE LOGITS
     prot
    0.30
    群
    0.27
     suck
    0.27
    éĢļ
    0.26
    Den
    0.26
    åį°
    0.26
    prot
    0.26
    ach
    0.25
    instr
    0.25
    program
    0.25
    Activations Density 0.119%

    No Known Activations