© 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. 32-TRANSCODER-HP
    7. 43690
    Prev
    Next
    INDEX
    Explanations

    Research and identification

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     identification
    -1.26
    ident
    -1.26
     ident
    -1.25
     Ident
    -1.21
    Ident
    -1.20
     identified
    -1.20
     identifying
    -1.17
     identify
    -1.16
    è¯ĨåĪ«
    -1.16
     identifies
    -1.12
    POSITIVE LOGITS
    峡
    0.27
    ubbo
    0.26
     TAS
    0.24
     hunting
    0.24
     Intersection
    0.23
    icals
    0.23
    uges
    0.23
    åĬŀæ¡Ī
    0.23
    umble
    0.23
    cré
    0.23
    Activations Density 0.002%

    No Known Activations