© 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. 130729
    Prev
    Next
    INDEX
    Explanations

    compete

    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
     mutual
    -0.45
    åĽ¢ç»ĵ
    -0.45
    äºĴ缸
    -0.44
    缸äºĴ
    -0.43
    å½¼æѤ
    -0.42
    åľ¨ä¸Ģèµ·
    -0.41
    åħ±åIJĮåĬªåĬĽ
    -0.41
    缸èģļ
    -0.40
     unity
    -0.40
     unified
    -0.40
    POSITIVE LOGITS
    shops
    0.27
    éŨå¤ĸ
    0.27
     CommandType
    0.25
    ableObject
    0.25
     infiltr
    0.24
     lifts
    0.24
    icio
    0.24
    éĵº
    0.24
    æīĢåľ¨
    0.24
    irable
    0.24
    Activations Density 36.407%

    No Known Activations