© 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. 31-TRANSCODER-HP
    7. 18294
    Prev
    Next
    INDEX
    Explanations

    programming code

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     localVar
    -0.28
     categor
    -0.25
    çħ¤æ°Ķ
    -0.25
    coal
    -0.24
    lust
    -0.24
    kle
    -0.24
    à¸ĸà¸Ļ
    -0.23
    ieved
    -0.23
     analy
    -0.23
    æ·ĭ
    -0.23
    POSITIVE LOGITS
    绪
    0.28
    éĹ»
    0.27
    abin
    0.26
    ateurs
    0.26
    ’av
    0.26
    mania
    0.25
    undra
    0.25
    aci
    0.25
    merc
    0.25
     prof
    0.25
    Activations Density 0.172%

    No Known Activations