© 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. 19367
    Prev
    Next
    INDEX
    Explanations

    and

    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
     ALLOW
    -0.29
    äºĨèµ·æĿ¥
    -0.28
    åħģ许
    -0.27
    allow
    -0.26
     Allowed
    -0.26
    åħģ
    -0.26
    onto
    -0.26
    ism
    -0.25
    大约
    -0.25
    约
    -0.25
    POSITIVE LOGITS
     thus
    0.63
    thus
    0.56
     Thus
    0.44
    Thus
    0.42
     can
    0.41
    è¿Ļæł·
    0.40
     must
    0.40
     asÃŃ
    0.37
    can
    0.37
    	can
    0.37
    Activations Density 1.245%

    No Known Activations