© 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. 18730
    Prev
    Next
    INDEX
    Explanations

    currently

    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
    %A
    -0.35
    èĦĬ
    -0.28
     spared
    -0.28
    %E
    -0.28
    iest
    -0.27
    %C
    -0.26
    ĥģ
    -0.26
    alty
    -0.25
    å°¬
    -0.25
    è¯ŀçĶŁ
    -0.25
    POSITIVE LOGITS
    èįī
    0.26
    wat
    0.25
     path
    0.25
    çĤ¹äºĨ
    0.25
    ounces
    0.25
    :path
    0.25
    path
    0.25
    andle
    0.24
    æĮĩ导æĦıè§ģ
    0.24
    å¤ĩæ¡Ī
    0.24
    Activations Density 0.224%

    No Known Activations