© 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. 30-TRANSCODER-HP
    7. 95086
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    æįķæįī
    -0.32
    æĮ£
    -0.30
    ãģ¤ãģij
    -0.30
    éļĶçĿĢ
    -0.28
     identification
    -0.28
    æĬĬèĩªå·±
    -0.27
     aud
    -0.26
     capturing
    -0.26
     maxx
    -0.26
     Overs
    -0.26
    POSITIVE LOGITS
     secrecy
    0.28
    ummy
    0.28
    åļı
    0.27
    羣çļĦ
    0.27
     really
    0.26
     suite
    0.26
    akis
    0.25
    åŃ©
    0.25
    Mahon
    0.25
     tucked
    0.25
    Activations Density 0.103%

    No Known Activations