© 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. 54826
    Prev
    Next
    INDEX
    Explanations

    Book/article references

    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
    uggy
    -0.27
    lus
    -0.26
     penetration
    -0.25
    tape
    -0.24
     ciné
    -0.24
    åįĥéĩĮ
    -0.24
     propensity
    -0.23
    utable
    -0.23
    ][-
    -0.23
     observational
    -0.23
    POSITIVE LOGITS
     ca
    0.47
    (ca
    0.27
    ca
    0.27
     c
    0.26
     Mode
    0.25
    qv
    0.25
     Co
    0.25
    tre
    0.24
     Ca
    0.24
     Vic
    0.24
    Activations Density 0.014%

    No Known Activations