© 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. 20-TRANSCODER-HP
    7. 107143
    Prev
    Next
    INDEX
    Explanations

    Conjunctions of contrast

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     sum
    -0.28
    åĪĨåĪ«
    -0.26
    鸯
    -0.26
     $(
    -0.25
     said
    -0.25
     ap
    -0.25
     плоÑīад
    -0.25
    ëĿ¼ëıĦ
    -0.25
    SS
    -0.24
    (ap
    -0.24
    POSITIVE LOGITS
    èľ£
    0.28
    åĴİ
    0.27
    éĢģä¸Ĭ
    0.27
    forth
    0.26
    hani
    0.25
     sponsorship
    0.25
    Submitting
    0.25
    ÑĢей
    0.24
    resizing
    0.24
    ImageContext
    0.24
    Activations Density 9.645%

    No Known Activations