© 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. 96139
    Prev
    Next
    INDEX
    Explanations

    ern

    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
     first
    -0.29
    -up
    -0.28
    up
    -0.28
    SW
    -0.28
    -over
    -0.28
    è¿ĩ
    -0.28
    åĪĩ
    -0.28
     Up
    -0.27
    -Up
    -0.27
    çļĦ第ä¸Ģ个
    -0.26
    POSITIVE LOGITS
    çĩİ
    0.29
    åħ¼ä»»
    0.28
     поÑĤ
    0.27
    è§Ĥ
    0.26
     jus
    0.26
    inders
    0.26
    ower
    0.26
    INDER
    0.25
    rior
    0.25
    romise
    0.25
    Activations Density 0.118%

    No Known Activations