© 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. 11-RES-JB
    11. 3649
    Prev
    Next
    INDEX
    Explanations

    the word "can't" or "can" when paired closely with certain verbs or topics

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "can't."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    pires
    -0.76
     Shrine
    -0.63
     Tablet
    -0.62
     Dickinson
    -0.62
     Pike
    -0.61
     Matters
    -0.57
     inspired
    -0.56
     Keen
    -0.56
     Roc
    -0.55
     spreads
    -0.54
    POSITIVE LOGITS
     afford
    1.20
     resist
    0.91
     tolerate
    0.87
     ignore
    0.86
    ¨
    0.84
     deny
    0.84
     conceive
    0.84
    Í
    0.82
     discriminate
    0.82
     underestimate
    0.82
    Activations Density 0.038%

    No Known Activations