© 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. 53097
    Prev
    Next
    INDEX
    Explanations

    say "out/in/from"

    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
    åĩº
    -0.42
     åĩ
    -0.30
    åĩºäºĨ
    -0.28
    åī§
    -0.27
     out
    -0.27
    å¡ij
    -0.26
    纳
    -0.26
    desc
    -0.25
    rett
    -0.25
     Aure
    -0.25
    POSITIVE LOGITS
    .from
    0.29
    (from
    0.28
     from
    0.27
    from
    0.27
    	from
    0.27
    -from
    0.26
    /delete
    0.26
    .free
    0.26
    /free
    0.25
    ured
    0.25
    Activations Density 0.005%

    No Known Activations