© 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. 31-TRANSCODER-HP
    7. 18468
    Prev
    Next
    INDEX
    Explanations

    colds and viruses

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ifiable
    -0.31
    igator
    -0.26
    .allocate
    -0.25
    åĩ¹
    -0.24
    -su
    -0.24
     Satellite
    -0.24
    edReader
    -0.24
    stasy
    -0.24
    å¹³åİŁ
    -0.24
    à¹Ģà¸Ĺà¸ŀ
    -0.23
    POSITIVE LOGITS
    GER
    0.30
     dr
    0.27
    agara
    0.27
     d
    0.26
    aghetti
    0.26
    rons
    0.26
    ugin
    0.25
    ogan
    0.25
    iral
    0.24
    age
    0.24
    Activations Density 1.192%

    No Known Activations