© 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. 11-RES-JB
    11. 10055
    Prev
    Next
    INDEX
    Explanations

    phrases containing the word "demeaning"

    oai_token-act-pair · gpt-3.5-turbo

    words and phrases related to derogatory or belittling commentary

    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
     unchecked
    -0.75
    $$$$
    -0.70
     Belt
    -0.66
     whiff
    -0.66
    ategory
    -0.62
     unexplained
    -0.62
     nose
    -0.61
    pmwiki
    -0.59
     heed
    -0.58
     [+
    -0.57
    POSITIVE LOGITS
    ufact
    1.05
    brance
    0.91
    culated
    0.91
    azon
    0.90
    oleon
    0.88
    acements
    0.87
    ogyn
    0.87
    rium
    0.86
    stration
    0.82
    andise
    0.78
    Activations Density 0.114%

    No Known Activations