© 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. 35-TRANSCODER-HP
    7. 39760
    Prev
    Next
    INDEX
    Explanations

    Trans

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     clo
    -0.29
     redund
    -0.27
     pilgrimage
    -0.26
     bufsize
    -0.26
     lane
    -0.26
     plateau
    -0.26
     purs
    -0.26
     palace
    -0.25
     quadr
    -0.25
     form
    -0.25
    POSITIVE LOGITS
    tim
    0.40
    tas
    0.38
    ted
    0.36
    tv
    0.36
    tm
    0.35
    tw
    0.35
    tg
    0.34
    t
    0.33
    cis
    0.33
    c
    0.32
    Activations Density 1.412%

    No Known Activations