© 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. 22-TRANSCODER-HP
    7. 88852
    Prev
    Next
    INDEX
    Explanations

    process

    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
    Negative Logits
    严
    -0.27
    <<<<<<<
    -0.25
    èĤ©
    -0.24
    ware
    -0.24
    ieber
    -0.24
    æĹĭ
    -0.24
    å¼Ĥ
    -0.24
    è¶Ĭ
    -0.24
    elay
    -0.23
     stagger
    -0.23
    POSITIVE LOGITS
    cido
    0.29
    iances
    0.25
     ($.
    0.24
    丰å¯Į
    0.23
    hic
    0.23
     Crescent
    0.23
    使人
    0.23
     thrust
    0.23
    è±IJå¯Į
    0.23
    ç¬ĶèĢħ
    0.23
    Activations Density 0.036%

    No Known Activations