© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 33-TRANSCODER-HP
    7. 52553
    Prev
    Next
    INDEX
    Explanations

    say "this"

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    æĽĻ
    -0.24
    urry
    -0.23
    itched
    -0.23
    quet
    -0.23
    åľ¨éĤ£éĩĮ
    -0.23
    éĤ£éĩĮ
    -0.22
     ÑĤам
    -0.21
    еждÑĥ
    -0.20
     esos
    -0.20
    quette
    -0.20
    POSITIVE LOGITS
     this
    1.61
    this
    1.59
    (this
    1.25
    	this
    1.22
    =this
    1.20
    -this
    1.19
    [this
    1.19
    æѤ
    1.13
    ,this
    1.13
    .this
    1.12
    Activations Density 0.315%

    No Known Activations