© 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. 11-RES-JB
    11. 9509
    Prev
    Next
    INDEX
    Explanations

    mentions of "dirty" or morally questionable actions/concepts

    oai_token-act-pair · gpt-3.5-turbo

    references to "dirty" or unethical practices and behaviors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    */(
    -1.20
    istically
    -0.82
    itech
    -0.80
    XT
    -0.78
    aic
    -0.77
    isol
    -0.76
    HCR
    -0.75
    ãĥĦ
    -0.74
    izations
    -0.73
    uther
    -0.73
    POSITIVE LOGITS
     laundry
    1.23
     tricks
    1.08
     linen
    1.06
     diapers
    0.97
     rotten
    0.93
     dirty
    0.92
     diaper
    0.88
     luc
    0.86
     trick
    0.85
     dishes
    0.83
    Activations Density 0.075%

    No Known Activations