© 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. 96169
    Prev
    Next
    INDEX
    Explanations

    rules

    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
     Tear
    -0.25
    液ä½ĵ
    -0.24
    主ä¸ļ
    -0.24
     Erf
    -0.24
     gaat
    -0.24
    æĶ¶
    -0.24
    ç»į
    -0.23
    èĪŀ
    -0.23
    ersh
    -0.23
    ceph
    -0.23
    POSITIVE LOGITS
    set
    0.45
     set
    0.38
    sets
    0.37
    -set
    0.36
    Set
    0.36
     sets
    0.33
     defining
    0.32
    SETS
    0.31
    ãĤ»ãĥĥãĥĪ
    0.31
    SET
    0.31
    Activations Density 0.045%

    No Known Activations