© 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. 10835
    Prev
    Next
    INDEX
    Explanations

    words related to employment positions or roles within an organization, particularly those at a higher level

    oai_token-act-pair · gpt-3.5-turbo

    references to staffers and their roles in various 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
     Bombay
    -0.69
    asca
    -0.67
    poses
    -0.65
     Sabha
    -0.64
    olithic
    -0.63
     Jub
    -0.63
    powers
    -0.62
    theless
    -0.62
    pha
    -0.61
    fb
    -0.61
    POSITIVE LOGITS
    irtual
    0.86
     staffers
    0.83
     staffer
    0.81
     Emails
    0.80
     reception
    0.77
     interns
    0.77
    email
    0.77
     stationed
    0.76
     tasked
    0.76
    hips
    0.75
    Activations Density 0.013%

    No Known Activations