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. 2-TRANSCODER-HP
    4. 6
    Prev
    Next
    INDEX
    Explanations

    code files/licenses

    np_max-act · gemini-2.0-flash

    occurrences of the character sequence “rest,” whether as a standalone word or embedded within other words or names.

    oai_token-act-pair · gpt-5Triggered by @liuliujiujiu05
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_2.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.2.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çľĭçĤ¹
    -0.27
    ukkit
    -0.27
    æµ´
    -0.26
    .nasa
    -0.26
    shoot
    -0.25
     Denied
    -0.25
    à¹Į
    -0.24
    اظ
    -0.24
    adays
    -0.24
     competing
    -0.24
    POSITIVE LOGITS
    ãĥªãĥ³
    0.28
    venida
    0.26
    ess
    0.26
    igest
    0.25
     Twice
    0.25
    Ãłn
    0.25
    åıĸå¾Ĺ
    0.25
    arti
    0.25
    iji
    0.25
    nid
    0.24
    Activations Density 43.776%

    No Known Activations