© 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. 0-RES-JB
    11. 10498
    Prev
    Next
    INDEX
    Explanations

    numbers representing amounts of money

    oai_token-act-pair · gpt-3.5-turbo

    references to numerical values, particularly those in the range of 800

    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
     spoiler
    -0.69
     Thy
    -0.68
     regression
    -0.65
     constit
    -0.64
     balancing
    -0.63
     Bram
    -0.62
    yll
    -0.62
     voting
    -0.60
     prediction
    -0.60
     lean
    -0.59
    POSITIVE LOGITS
    800
    3.31
     800
    1.99
    877
    1.83
    888
    1.83
    8000
    1.82
    700
    1.80
    900
    1.76
    866
    1.76
    600
    1.61
    400
    1.61
    Activations Density 0.014%

    No Known Activations