© 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. 25-TRANSCODER-HP
    7. 108853
    Prev
    Next
    INDEX
    Explanations

    Article text

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    idian
    -0.33
     testcase
    -0.28
    -issue
    -0.28
     vitae
    -0.25
    TestCase
    -0.25
    çľĭå¾Ĺ
    -0.25
    ulaire
    -0.24
    iliar
    -0.24
    KIT
    -0.24
    ertools
    -0.24
    POSITIVE LOGITS
    During
    0.28
    á»Ĩ
    0.27
     During
    0.27
     during
    0.25
     sys
    0.24
     gu
    0.24
    èĩªçĦ¶æĺ¯
    0.24
    eps
    0.23
    .IsMatch
    0.23
    AttributeValue
    0.23
    Activations Density 1.049%

    No Known Activations