© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 18713
    Prev
    Next
    INDEX
    Explanations

    Code snippets

    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
    Not in Any Lists

    No Comments

    Negative Logits
    conscious
    -0.32
    èĤ²
    -0.29
     Asc
    -0.27
    æĸij
    -0.26
    noc
    -0.26
    asil
    -0.26
    opo
    -0.26
    æĪIJåĵģ
    -0.25
     conscious
    -0.25
    æĬ¼
    -0.25
    POSITIVE LOGITS
    oring
    0.30
     redd
    0.28
    ç½ķ
    0.27
    -red
    0.26
    ole
    0.25
     communicating
    0.25
    meer
    0.24
    æĬķèµĦåŁºéĩij
    0.24
    çļĦéĴ±
    0.24
    redni
    0.24
    Activations Density 0.009%

    No Known Activations