© 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. 53507
    Prev
    Next
    INDEX
    Explanations

    locations

    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
     charg
    -0.28
    raison
    -0.28
    aus
    -0.28
    psz
    -0.26
    vic
    -0.26
     chronic
    -0.26
     charge
    -0.26
    دÙĦ
    -0.26
    uded
    -0.26
    ä»İä¸Ń
    -0.26
    POSITIVE LOGITS
    vol
    0.27
    失
    0.26
     Hier
    0.24
     Jefferson
    0.24
    etto
    0.23
    æ±Ĥ
    0.23
     cert
    0.23
    Hier
    0.23
    è°¤
    0.23
    éĻ°
    0.23
    Activations Density 0.011%

    No Known Activations