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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 2752
    Prev
    Next
    INDEX
    Explanations

    words related to exceptions in a context, especially in programming or law

    oai_token-act-pair · gpt-3.5-turbo

    mentions of exceptions or deviations from rules or norms

    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
    iled
    -0.71
    weet
    -0.65
    athy
    -0.61
    Â
    -0.59
    uve
    -0.58
     began
    -0.58
    walker
    -0.57
    rat
    -0.56
     chant
    -0.56
    air
    -0.56
    POSITIVE LOGITS
     exceptions
    4.23
     exception
    2.27
     exemptions
    1.93
     caveats
    1.80
     Exception
    1.57
     loopholes
    1.49
     deviations
    1.46
     examples
    1.32
    ceptions
    1.30
     distinctions
    1.29
    Activations Density 0.007%

    No Known Activations