© 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. 25-TRANSCODER-HP
    7. 108439
    Prev
    Next
    INDEX
    Explanations

    technical language

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    лав
    -0.31
    åĪ©äºļ
    -0.27
    ä»Ļ
    -0.27
    frauen
    -0.26
     Collector
    -0.26
    ç¨³åĽº
    -0.26
    minster
    -0.25
    è¿ľè¿ľ
    -0.25
    legg
    -0.25
    ·¨
    -0.25
    POSITIVE LOGITS
    ObjectContext
    0.34
    -fit
    0.27
    Fit
    0.27
    us
    0.26
    ua
    0.26
    mo
    0.25
    duc
    0.24
     fit
    0.24
     Fit
    0.24
    èĦ±
    0.24
    Activations Density 0.924%

    No Known Activations