© 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. 18903
    Prev
    Next
    INDEX
    Explanations

    names/identifiers

    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
     Lips
    -0.27
     exp
    -0.27
     kald
    -0.27
    åı²è¯Ĺ
    -0.26
    olie
    -0.26
    ancybox
    -0.26
     kaz
    -0.26
    ipsis
    -0.25
    regexp
    -0.25
     Jesse
    -0.25
    POSITIVE LOGITS
    æĮģ
    0.28
    çĵ¦
    0.27
     hemorrh
    0.26
    bones
    0.26
     Cup
    0.25
    äºĨäºĽ
    0.25
    ament
    0.25
     Hab
    0.25
    YNAMIC
    0.25
     Soul
    0.25
    Activations Density 0.102%

    No Known Activations