© 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. 18938
    Prev
    Next
    INDEX
    Explanations

    scientists

    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
    åĩ¸æĺ¾
    -0.26
    âĸ¾
    -0.25
     contributors
    -0.24
    silent
    -0.24
    åĩºåĶ®
    -0.24
    erg
    -0.24
    ä¿Ŀæ´ģ
    -0.23
    强åĮĸ
    -0.23
    ãĤĴè¦ĭãģĽ
    -0.23
    æĬķ稿
    -0.23
    POSITIVE LOGITS
     once
    0.39
    éĥ½ä¸įèĥ½
    0.35
    Called
    0.35
     spec
    0.33
    once
    0.33
     believed
    0.32
     arrived
    0.32
    çĽ¸ä¿¡
    0.32
     discovered
    0.32
    éĥ½è®¤ä¸º
    0.31
    Activations Density 0.032%

    No Known Activations