© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 794
    Prev
    Next
    INDEX
    Explanations

    words related to curses or expressing strong negative emotions

    oai_token-act-pair · gpt-3.5-turbo

    occurrences and variations of the word "curse."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    姫
    -0.92
    nington
    -0.84
    visors
    -0.80
    arger
    -0.75
    arnaev
    -0.74
    ighters
    -0.74
    anooga
    -0.73
    ullivan
    -0.73
    olitan
    -0.72
    endment
    -0.72
    POSITIVE LOGITS
     curse
    1.07
     curses
    0.93
     cursing
    0.85
     cursed
    0.85
     Curse
    0.76
    hammer
    0.70
     fate
    0.70
     inflicted
    0.69
    breakers
    0.69
     wrench
    0.68
    Activations Density 0.010%

    No Known Activations