© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 64646
    Prev
    Next
    INDEX
    Explanations

    method

    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
    Negative Logits
    agit
    -0.27
     Hash
    -0.26
    LAG
    -0.25
    audi
    -0.24
    ax
    -0.24
    akk
    -0.24
    urable
    -0.24
    atable
    -0.24
    ashi
    -0.24
    Hash
    -0.23
    POSITIVE LOGITS
    -io
    0.26
     benefits
    0.26
    æ¶īåıĬ
    0.25
    çļĦ好å¤Ħ
    0.25
    ÑĨии
    0.25
    ä»Ŀ
    0.25
    ä¸İåIJ¦
    0.24
    æĺ¯åIJ¦
    0.24
    _INS
    0.24
    çī©
    0.24
    Activations Density 11.906%

    No Known Activations