© 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. 7-RES-JB
    11. 2742
    Prev
    Next
    INDEX
    Explanations

    numerical values indicating an increase or addition

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "additional" and similar phrases indicating increments or extra amounts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bows
    -0.94
    lees
    -0.86
    casts
    -0.82
    encers
    -0.79
    images
    -0.78
    tics
    -0.78
    sites
    -0.75
    masters
    -0.74
     treaties
    -0.73
    events
    -0.72
    POSITIVE LOGITS
     layer
    1.19
     hurdle
    1.03
     step
    1.01
     handful
    0.97
     piece
    0.95
     complication
    0.93
     caveat
    0.92
     couple
    0.92
     glimpse
    0.91
     decade
    0.91
    Activations Density 0.072%

    No Known Activations