© 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. 53151
    Prev
    Next
    INDEX
    Explanations

    say "statement"

    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
    clause
    -0.28
    åı°è¯į
    -0.26
    eneg
    -0.26
     Wort
    -0.26
    prite
    -0.26
     cümle
    -0.26
     Clause
    -0.26
    prites
    -0.26
    scope
    -0.26
    壮
    -0.25
    POSITIVE LOGITS
     rest
    0.48
    Rest
    0.39
    rest
    0.38
     Rest
    0.37
    _rest
    0.32
     heading
    0.30
     st
    0.29
     resta
    0.28
    -rest
    0.28
     resto
    0.27
    Activations Density 0.003%

    No Known Activations