© 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. 31-TRANSCODER-HP
    7. 19316
    Prev
    Next
    INDEX
    Explanations

    stability

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ë²
    -0.27
    çĥĽ
    -0.27
    à¹Ĥม
    -0.27
    кова
    -0.25
    serter
    -0.25
     enumerable
    -0.25
    åĭĺå¯Ł
    -0.25
    esk
    -0.24
    ophobic
    -0.24
     sta
    -0.24
    POSITIVE LOGITS
    s
    0.77
    sist
    0.38
    sand
    0.37
    sphere
    0.37
    sing
    0.35
    sheets
    0.34
    sip
    0.34
    sport
    0.34
    sav
    0.34
    sense
    0.31
    Activations Density 0.027%

    No Known Activations