Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 6954
    Prev
    Next
    INDEX
    Explanations

    words related to protection or safeguarding

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "protecting" and its variations in relation to safety and security

    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
    rose
    -0.78
    izzle
    -0.73
    Joy
    -0.66
     Quarterly
    -0.61
    Sing
    -0.61
     Ket
    -0.60
    iga
    -0.60
    iston
    -0.60
    zig
    -0.60
    ude
    -0.59
    POSITIVE LOGITS
     protecting
    3.36
    protect
    2.37
     guarding
    2.16
     preventing
    2.08
     protect
    2.05
     safegu
    2.02
     shielding
    1.96
     preserving
    1.95
     Protect
    1.94
    Protect
    1.93
    Activations Density 0.023%

    No Known Activations