© 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. 4-RES-JB
    11. 23682
    Prev
    Next
    INDEX
    Explanations

    the name "Nixon"

    oai_token-act-pair · gpt-3.5-turbo

    references to Richard Nixon and related political events

    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
    medium
    -0.71
     Flavoring
    -0.70
    optional
    -0.69
    topic
    -0.68
    FORM
    -0.68
    malink
    -0.67
    tered
    -0.67
    ================================================================
    -0.66
    acters
    -0.65
    essional
    -0.63
    POSITIVE LOGITS
    shire
    1.02
     Nixon
    0.94
    iere
    0.70
     Jr
    0.68
    oleon
    0.67
     Kissinger
    0.67
    enegger
    0.66
     Sut
    0.65
     Jong
    0.65
     submar
    0.64
    Activations Density 0.008%

    No Known Activations