© 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. 18479
    Prev
    Next
    INDEX
    Explanations

    University addresses

    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
    ä¸įæĺİ
    -0.29
     Dems
    -0.29
     math
    -0.25
     presidents
    -0.25
     pres
    -0.25
     differential
    -0.25
    Dig
    -0.25
     governmental
    -0.24
    iete
    -0.24
    lections
    -0.23
    POSITIVE LOGITS
    åĽŀ家
    0.27
    çĽĴ
    0.25
    anger
    0.25
    æŁ¥å¤Ħ
    0.24
    steen
    0.24
    ENA
    0.24
    home
    0.24
    åįķ车
    0.24
    æĶ¾åĪ°
    0.23
     Boh
    0.23
    Activations Density 0.136%

    No Known Activations