© 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. 54029
    Prev
    Next
    INDEX
    Explanations

    Programming code

    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
     softened
    -0.32
    rede
    -0.29
    Uniform
    -0.29
     Uniform
    -0.28
     lodged
    -0.27
     bowed
    -0.27
    æĺİçıł
    -0.27
    çĶ¨é¤IJ
    -0.26
    浪费
    -0.26
    зав
    -0.26
    POSITIVE LOGITS
    ek
    0.35
    ea
    0.33
    ivirus
    0.32
    ec
    0.31
    ean
    0.31
    htub
    0.31
    ef
    0.31
    ee
    0.30
    eq
    0.30
    enen
    0.29
    Activations Density 0.006%

    No Known Activations