© 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. 4-RES-JB
    11. 20105
    Prev
    Next
    INDEX
    Explanations

    phrases with a focus on individual entities or organizations

    oai_token-act-pair · gpt-3.5-turbo

    references to individual cases or instances within a broader context

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

    No Comments

    Negative Logits
     Cance
    -0.75
     Newsletter
    -0.70
     Compliance
    -0.67
    wolves
    -0.67
     Lor
    -0.67
    wives
    -0.64
     Ary
    -0.63
     Lawyers
    -0.63
    humans
    -0.63
    yrights
    -0.61
    POSITIVE LOGITS
     individually
    1.15
     separately
    0.95
     mango
    0.83
     respective
    0.77
     equally
    0.75
     independently
    0.74
    ieth
    0.72
     facet
    0.72
     sexes
    0.70
     increment
    0.69
    Activations Density 0.139%

    No Known Activations