© 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. 32-TRANSCODER-HP
    7. 42132
    Prev
    Next
    INDEX
    Explanations

    say "capitalized word"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     dap
    -0.25
     deltaTime
    -0.24
    ellar
    -0.24
    éĹ®è´£
    -0.23
    amina
    -0.23
    èľľèľĤ
    -0.23
     setFrame
    -0.23
    åħħåĪĨåĪ©çĶ¨
    -0.23
    éĥ½è¦ģ
    -0.23
    quil
    -0.23
    POSITIVE LOGITS
    åħ¨å±Ģ
    0.32
    edis
    0.29
    ITH
    0.26
    Univers
    0.26
     intellectually
    0.26
    .ge
    0.26
    ire
    0.25
     emoc
    0.25
     ste
    0.25
     Lith
    0.25
    Activations Density 0.001%

    No Known Activations