© 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. 89226
    Prev
    Next
    INDEX
    Explanations

    lightning

    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
    è¿ĻäºĽéĹ®é¢ĺ
    -0.30
    èī¯
    -0.26
    -lfs
    -0.25
    çĦ¶æĺ¯
    -0.25
    éal
    -0.25
     çöz
    -0.25
    Trait
    -0.25
    _solver
    -0.25
    MemoryWarning
    -0.24
    /from
    -0.24
    POSITIVE LOGITS
    eme
    0.28
    dig
    0.25
    utow
    0.24
    é¡¿
    0.24
    æĮ¤
    0.23
    åıijæİĺ
    0.23
    oon
    0.23
    æ´¿
    0.23
     bord
    0.23
    ç¬Ķ
    0.23
    Activations Density 10.141%

    No Known Activations