Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Qwen3-4B
    3. 23-TRANSCODER-HP
    4. 15013
    Prev
    Next
    INDEX
    Explanations

    Writing or arriving

    np_max-act · gemini-2.0-flash

    verbs indicating concrete actions taken by someone (often the author) to do, create, or try something, especially in technical/problem‑solving contexts.

    oai_token-act-pair · gpt-5Triggered by @liuliujiujiu05
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_23.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.23.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     recently
    -0.28
    åĩºèĩª
    -0.27
    Recently
    -0.26
    .TryParse
    -0.25
     rookies
    -0.25
    è§ģè¯Ĩ
    -0.25
    _recent
    -0.24
    psc
    -0.24
     counts
    -0.24
    çĮ¹
    -0.23
    POSITIVE LOGITS
    ä¸Ģæľ¬ä¹¦
    0.29
     lap
    0.29
    æĮĤ
    0.28
    åģľæŃ¢
    0.27
    éģĩä¸Ĭ
    0.26
    两çĤ¹
    0.26
    以ä¸ĭ
    0.26
    habit
    0.25
    fld
    0.24
    ä¸įå¼Ģ
    0.24
    Activations Density 0.021%

    No Known Activations