© 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. 2844
    Prev
    Next
    INDEX
    Explanations

    economic and financial terms or indicators

    oai_token-act-pair · gpt-3.5-turbo

    keywords related to economic data and trends

    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
     Scientology
    -0.76
    authent
    -0.67
     invasive
    -0.65
     pedoph
    -0.65
    UTH
    -0.65
     Lovecraft
    -0.64
    Spons
    -0.64
     copyrighted
    -0.64
     Canaver
    -0.64
     ethical
    -0.63
    POSITIVE LOGITS
     forecasts
    1.07
    flation
    1.05
     slump
    0.99
     inflation
    0.97
     forecast
    0.96
     recession
    0.95
     GDP
    0.93
     slowdown
    0.92
     forecasting
    0.91
     outlook
    0.89
    Activations Density 0.307%

    No Known Activations