© 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. 53689
    Prev
    Next
    INDEX
    Explanations

    Negative events/situations

    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
     dances
    -0.40
     conflicts
    -0.37
     accidents
    -0.35
     explosions
    -0.34
    errors
    -0.34
     crashes
    -0.34
     battles
    -0.33
     wars
    -0.33
     fires
    -0.32
     fights
    -0.32
    POSITIVE LOGITS
     infection
    0.33
     problem
    0.33
     incident
    0.32
     complaint
    0.30
    pute
    0.30
     issue
    0.30
     altercation
    0.29
     question
    0.28
    æ¡Ī
    0.28
     episode
    0.28
    Activations Density 0.035%

    No Known Activations