© 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. 1-RES-JB
    11. 3849
    Prev
    Next
    INDEX
    Explanations

    words related to exploiting legal or regulatory loopholes

    oai_token-act-pair · gpt-3.5-turbo

    references to legal loopholes and ways to exploit them

    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
    semble
    -0.70
    oran
    -0.69
    grave
    -0.69
     Kinnikuman
    -0.68
    haps
    -0.66
    yss
    -0.66
     pastoral
    -0.65
    Ķ
    -0.63
     ser
    -0.61
    image
    -0.60
    POSITIVE LOGITS
     loopholes
    1.55
     loophole
    1.46
    glers
    0.98
     exemptions
    0.93
     deductions
    0.84
    witz
    0.81
     circumvent
    0.80
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    0.80
     cheat
    0.75
     backdoor
    0.72
    Activations Density 0.009%

    No Known Activations