© 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. 18628
    Prev
    Next
    INDEX
    Explanations

    Non-English words

    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
    èĤ®
    -0.27
     mist
    -0.27
    NET
    -0.26
    èĺ¸
    -0.26
    OP
    -0.26
    iction
    -0.26
    opa
    -0.26
    oph
    -0.26
    adamente
    -0.25
    Dr
    -0.25
    POSITIVE LOGITS
    弧
    0.29
    线ä¸ĭ
    0.29
    ALER
    0.28
    ÑĤин
    0.27
    lines
    0.27
    ALAR
    0.26
    åıī
    0.26
     strains
    0.25
    ä¸ĭè¡Į
    0.25
    åıijå±ķ缮æłĩ
    0.25
    Activations Density 0.175%

    No Known Activations