© 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. 30-TRANSCODER-HP
    7. 96459
    Prev
    Next
    INDEX
    Explanations

    magazine

    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
    Not in Any Lists

    No Comments

    Negative Logits
    åĽłåľ°
    -0.32
    åIJ¬äºĨ
    -0.26
    ders
    -0.26
     Forum
    -0.26
    æ´Ĺæīĭ
    -0.25
    isch
    -0.25
    çĹ£
    -0.25
    æİĴè¡Į
    -0.25
    èİĵ
    -0.24
    åIJ¬åĬĽ
    -0.24
    POSITIVE LOGITS
     cover
    0.40
    -cover
    0.39
     dedicated
    0.39
     issue
    0.36
     dedic
    0.36
     published
    0.36
    cover
    0.35
     covering
    0.34
     devoted
    0.34
     publish
    0.33
    Activations Density 0.125%

    No Known Activations