© 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. 13940
    Prev
    Next
    INDEX
    Explanations

    phrases related to the Indian city of Madras or terms containing "ras"

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of the term "ras."

    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
     PC
    -0.66
     payroll
    -0.65
     account
    -0.64
     stamp
    -0.64
     check
    -0.62
     checks
    -0.62
     testing
    -0.62
     checking
    -0.61
     NC
    -0.60
     disarm
    -0.59
    POSITIVE LOGITS
    ras
    4.54
    ra
    1.61
    ran
    1.45
    rus
    1.43
    ros
    1.42
    ram
    1.40
    rase
    1.40
    rals
    1.37
    ral
    1.36
    rad
    1.33
    Activations Density 0.011%

    No Known Activations