© 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. 0-RES-JB
    11. 16940
    Prev
    Next
    INDEX
    Explanations

    mentions of steel-related terms like "steel maker," "stainless steel," and "steel products."

    oai_token-act-pair · gpt-3.5-turbo

    references to steel and its various forms and uses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Samar
    -0.73
     Trends
    -0.72
     Parent
    -0.71
    mess
    -0.71
    Mess
    -0.70
    heter
    -0.70
    Events
    -0.66
     Kardash
    -0.65
     Parties
    -0.65
    Mp
    -0.64
    POSITIVE LOGITS
     steel
    3.60
    steel
    2.66
     Steel
    2.46
    Steel
    2.24
     aluminium
    1.61
     metal
    1.61
     stainless
    1.60
     iron
    1.57
     aluminum
    1.57
     titanium
    1.54
    Activations Density 0.007%

    No Known Activations