© 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. 8123
    Prev
    Next
    INDEX
    Explanations

    names of individuals, potentially associated with movies or other works, and medical terms like "coronary heart disease" and "Atkins-like diets."

    oai_token-act-pair · gpt-3.5-turbo

    mentions of notable individuals and their works

    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
     consulate
    -0.75
     IMAGES
    -0.72
     mole
    -0.72
     customs
    -0.70
     vet
    -0.69
     nick
    -0.69
    geries
    -0.68
     counterfe
    -0.68
     ransom
    -0.68
     Fasc
    -0.68
    POSITIVE LOGITS
     Lean
    2.89
    Lean
    2.07
     Atkins
    1.95
    assi
    1.29
    ÃŁ
    1.20
     Garc
    1.17
    grid
    1.13
     Ung
    1.03
     Manson
    1.02
    rane
    1.00
    Activations Density 0.042%

    No Known Activations