© 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. 22-TRANSCODER-HP
    7. 111507
    Prev
    Next
    INDEX
    Explanations

    code snippets

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .yy
    -0.29
    ingular
    -0.26
    flt
    -0.25
    hma
    -0.25
    ipa
    -0.25
     lyon
    -0.25
    young
    -0.25
    æľĢåŁºæľ¬
    -0.25
    ustral
    -0.24
     leaps
    -0.24
    POSITIVE LOGITS
     dislikes
    0.33
     оÑĤказ
    0.32
    æĭĴç»Ŀ
    0.32
     dislike
    0.31
    æĭĴ
    0.31
     refusal
    0.31
     rejected
    0.31
     insufficient
    0.31
     fit
    0.31
     inability
    0.30
    Activations Density 0.014%

    No Known Activations