Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 4670
    Prev
    Next
    INDEX
    Explanations

    words related to fatal incidents or tragedies

    oai_token-act-pair · gpt-3.5-turbo

    instances of fatal incidents or deaths

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ESS
    -0.63
     Expansion
    -0.63
     Tee
    -0.62
    ãĢģ
    -0.61
     sits
    -0.61
    oodle
    -0.61
     Painter
    -0.61
    AMP
    -0.61
    Row
    -0.60
     enjoys
    -0.59
    POSITIVE LOGITS
     fatal
    3.71
     Fatal
    1.99
     deadly
    1.94
     fatally
    1.93
     lethal
    1.90
     fatalities
    1.44
     tragic
    1.42
     catastrophic
    1.39
     deaths
    1.38
     disastrous
    1.33
    Activations Density 0.020%

    No Known Activations