© 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. 1-RES-JB
    11. 23340
    Prev
    Next
    INDEX
    Explanations

    medical units and measurements, such as doses in mg and dB

    oai_token-act-pair · gpt-3.5-turbo

    measurements and quantities related to substances and their concentrations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    locked
    -0.76
    stru
    -0.75
    raged
    -0.70
     Brav
    -0.69
    Row
    -0.68
    bro
    -0.68
     burg
    -0.67
    broken
    -0.66
    pol
    -0.66
    Forge
    -0.65
    POSITIVE LOGITS
     Ago
    0.95
    gans
    0.88
    uana
    0.78
    aceutical
    0.78
    imeters
    0.77
     EDT
    0.77
     sulf
    0.75
    ï¸
    0.74
     Seym
    0.73
    emonic
    0.73
    Activations Density 0.015%

    No Known Activations