© 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. 64178
    Prev
    Next
    INDEX
    Explanations

    Complete sentences

    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
    çī¢è®°ä½¿åij½
    -0.28
     McD
    -0.28
    åĬ£
    -0.27
    o
    -0.27
    ter
    -0.26
    ue
    -0.26
    urgeon
    -0.26
    toupper
    -0.26
    åīĮ
    -0.26
    oi
    -0.25
    POSITIVE LOGITS
     Enough
    0.26
    ihn
    0.26
    mint
    0.25
    athi
    0.24
     glimpse
    0.24
     Wich
    0.24
    åĮºåĨħ
    0.24
     },↵↵↵
    0.23
    åĩłæ¬¡
    0.23
    mát
    0.23
    Activations Density 8.671%

    No Known Activations