© 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. 108324
    Prev
    Next
    INDEX
    Explanations

    "one" or single

    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
    upp
    -0.29
    éĢł
    -0.29
    umbo
    -0.26
     inaccessible
    -0.25
     inade
    -0.25
    acity
    -0.25
    æIJŃéħį
    -0.25
     aw
    -0.25
    çĶļ
    -0.25
     Jaune
    -0.24
    POSITIVE LOGITS
    æĹłéľĢ
    0.44
    ç®ĢåĮĸ
    0.43
     simpl
    0.41
     simplify
    0.39
    åħį
    0.36
     streamline
    0.35
     simplified
    0.35
    ä¸įå¿ħ
    0.35
    ä¸įéľĢè¦ģ
    0.34
     eliminates
    0.34
    Activations Density 75.322%

    No Known Activations