© 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. 22-TRANSCODER-HP
    7. 97648
    Prev
    Next
    INDEX
    Explanations

    Code and legal text

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    parm
    -0.24
    çĶ·åıĭ
    -0.23
     commissioner
    -0.23
    antage
    -0.23
     getChild
    -0.22
    roker
    -0.22
     koÅŁ
    -0.22
    ä¾
    -0.22
    ubbles
    -0.22
     Banc
    -0.22
    POSITIVE LOGITS
    "?
    0.25
     pist
    0.25
    way
    0.24
    key
    0.24
    [*
    0.24
    å¾Ĺ好
    0.24
    "+
    0.23
     Mit
    0.23
     pains
    0.23
     guten
    0.23
    Activations Density 0.041%

    No Known Activations