© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 4-RES-JB
    11. 7402
    Prev
    Next
    INDEX
    Explanations

    phrases related to avoiding involvement or participation in various activities or situations

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the phrase "stay out."

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

    No Comments

    Negative Logits
    BILITY
    -0.78
     arsen
    -0.71
    Export
    -0.70
     etched
    -0.68
     laureate
    -0.65
     fabrication
    -0.63
    cius
    -0.61
    matically
    -0.60
    Edited
    -0.59
     titanium
    -0.59
    POSITIVE LOGITS
    posts
    0.99
    doors
    0.95
    skirts
    0.90
    fitted
    0.88
    stretched
    0.88
    door
    0.85
    bur
    0.80
    ouk
    0.79
    casts
    0.75
    stay
    0.75
    Activations Density 0.055%

    No Known Activations