© 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. 20069
    Prev
    Next
    INDEX
    Explanations

    instances of the word "protect" in contexts related to ensuring safety or security

    oai_token-act-pair · gpt-3.5-turbo

    contexts related to the theme of protection

    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
    hour
    -0.77
    VP
    -0.74
    Hour
    -0.70
    ffer
    -0.69
    ellen
    -0.65
     thence
    -0.64
     Weeks
    -0.64
    LINE
    -0.64
    dash
    -0.64
    ching
    -0.63
    POSITIVE LOGITS
     protecting
    1.12
    orate
    1.06
    ively
    1.04
     protect
    0.99
     protects
    0.98
     Protect
    0.96
     protection
    0.96
    rets
    0.91
     shields
    0.89
     shielding
    0.89
    Activations Density 0.019%

    No Known Activations