© 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. 53678
    Prev
    Next
    INDEX
    Explanations

    documents, standards and errors

    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
    amide
    -0.27
     fibre
    -0.26
    Clause
    -0.26
     behaviours
    -0.26
     licence
    -0.25
    èĽĭ
    -0.25
    fiber
    -0.25
    Facade
    -0.25
    èĢĨ
    -0.24
    oles
    -0.24
    POSITIVE LOGITS
     Address
    0.37
    address
    0.35
     movement
    0.34
     addressed
    0.34
     address
    0.34
    åľ°åĿĢ
    0.34
     addresses
    0.34
     mov
    0.32
     addressing
    0.32
    -address
    0.32
    Activations Density 0.006%

    No Known Activations