© 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. 18-TRANSCODER-HP
    7. 88906
    Prev
    Next
    INDEX
    Explanations

    unusual textual content

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æĤĴ
    -0.28
    åıĭæĥħ
    -0.26
    ijľ
    -0.25
    untos
    -0.25
     contrad
    -0.25
    uctions
    -0.24
    ahir
    -0.24
    黼
    -0.24
    extr
    -0.23
     Bis
    -0.23
    POSITIVE LOGITS
    ç¼Ģ
    0.27
    JOR
    0.27
    /opt
    0.27
    å°±æĺ¯åľ¨
    0.27
    ä¸īåĪĨ
    0.26
    å°±åºĶ该
    0.25
    CUSTOM
    0.25
    owa
    0.24
    ä¼į
    0.24
    jian
    0.24
    Activations Density 0.207%

    No Known Activations