© 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. 33-TRANSCODER-HP
    7. 53600
    Prev
    Next
    INDEX
    Explanations

    Alexander

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Morm
    -0.27
     Terr
    -0.25
    ERY
    -0.25
    osition
    -0.25
    æĽ©
    -0.25
    pos
    -0.25
    +len
    -0.24
    ä¸Ģé¡¿
    -0.24
     Hamp
    -0.24
     newPos
    -0.24
    POSITIVE LOGITS
    otron
    0.28
     ÐŁÑĢед
    0.27
    Vo
    0.27
     outlier
    0.27
     rip
    0.27
    è¯į
    0.26
    夹
    0.26
    æ®ī
    0.26
    Ļ¨
    0.26
     mer
    0.25
    Activations Density 0.010%

    No Known Activations