© 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. 30-TRANSCODER-HP
    7. 96222
    Prev
    Next
    INDEX
    Explanations

    9

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ört
    -0.29
    .observe
    -0.26
    .launch
    -0.26
     launch
    -0.26
    pt
    -0.26
    quote
    -0.25
    çĶ¸
    -0.25
     Built
    -0.24
     ore
    -0.24
     cliff
    -0.24
    POSITIVE LOGITS
    (ra
    0.28
    itational
    0.28
    Ñģо
    0.27
    åįĩ
    0.26
    phia
    0.25
    iona
    0.25
     vac
    0.25
    åįĩéĻį
    0.25
     Seventh
    0.24
    èĢį
    0.24
    Activations Density 0.010%

    No Known Activations