© 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. 23258
    Prev
    Next
    INDEX
    Explanations

    a specific text string, likely a file name or identifier such as "Forencich" or "png."

    oai_token-act-pair · gpt-3.5-turbo

    specific references to images and notable entities or topics, particularly in a digital context

    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
    Age
    -0.78
     Horizon
    -0.72
    ears
    -0.71
     dealership
    -0.71
    peak
    -0.69
     hom
    -0.69
     Motors
    -0.67
     Settlement
    -0.67
    service
    -0.67
     bunk
    -0.67
    POSITIVE LOGITS
    nc
    2.94
    png
    1.86
     PNG
    1.62
     Orb
    1.09
     proportional
    1.03
    ombo
    0.97
     orb
    0.95
     Abram
    0.93
    ĸļ
    0.92
    zag
    0.92
    Activations Density 0.053%

    No Known Activations