© 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. 8553
    Prev
    Next
    INDEX
    Explanations

    terms related to various types of roles or positions, including players, borrowers, commuters, consumers, investors, lawmakers, cities, shareholders, riders, governments, companies, people, and mothers

    oai_token-act-pair · gpt-3.5-turbo

    terms related to various groups or categories of people involved in economic or social contexts

    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
    REDACTED
    -0.60
    uran
    -0.58
    Male
    -0.56
    Shock
    -0.56
    YL
    -0.56
    ;;;;;;;;;;;;
    -0.55
    family
    -0.55
    ige
    -0.54
    SL
    -0.54
    atan
    -0.53
    POSITIVE LOGITS
     forgot
    0.72
     cannot
    0.63
     invent
    0.60
     learned
    0.59
     innov
    0.59
     weren
    0.58
     programmers
    0.58
     reverted
    0.58
     forg
    0.57
     migrated
    0.57
    Activations Density 0.430%

    No Known Activations