© 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. 33-TRANSCODER-HP
    7. 53715
    Prev
    Next
    INDEX
    Explanations

    Narrative text

    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
    Not in Any Lists

    No Comments

    Negative Logits
     happens
    -0.29
     happening
    -0.27
     happen
    -0.27
    èĥ¡èIJĿåįľ
    -0.23
    TemplateName
    -0.23
     stationary
    -0.23
    Cx
    -0.23
    Gravity
    -0.23
    uctions
    -0.23
    åľī
    -0.23
    POSITIVE LOGITS
     lay
    0.40
    lay
    0.32
     Lay
    0.31
     pie
    0.26
     curs
    0.26
    lost
    0.26
     paced
    0.26
    ëĿ¼
    0.25
    =yes
    0.25
     lays
    0.25
    Activations Density 0.079%

    No Known Activations