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

    Neuronpedia

    APICircuit 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. 1-RES-JB
    6. 18611
    Prev
    Next
    INDEX
    Explanations

    instances of event cancellations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "canceled" or "cancelled."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Inst
    -0.83
    adr
    -0.79
    Appearance
    -0.79
    bour
    -0.77
    advertisement
    -0.71
    MAC
    -0.71
    METHOD
    -0.69
    ker
    -0.65
    Honest
    -0.65
    Foot
    -0.65
    POSITIVE LOGITS
     cancell
    1.32
     cancel
    1.20
     cance
    1.17
     cancelled
    1.15
     canceled
    1.12
     cancellation
    1.02
     screenings
    0.86
     refunds
    0.81
     postponed
    0.76
     abruptly
    0.75
    Activations Density 0.008%

    No Known Activations