© 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. 18799
    Prev
    Next
    INDEX
    Explanations

    phrases related to outlining or describing a plan or situation

    oai_token-act-pair · gpt-3.5-turbo

    phrases that describe laying out or outlining details and cases

    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
    IUM
    -0.67
    lez
    -0.66
    iversary
    -0.65
    AUT
    -0.65
    EEP
    -0.63
    ãĥĺãĥ©
    -0.63
    taboola
    -0.62
    assic
    -0.62
    asion
    -0.61
    enf
    -0.61
    POSITIVE LOGITS
     how
    1.37
     why
    1.22
     exactly
    1.02
     strategies
    0.99
    why
    0.97
     what
    0.96
     ways
    0.94
     myriad
    0.90
     WHY
    0.89
     steps
    0.89
    Activations Density 0.181%

    No Known Activations