© 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. 4-RES-JB
    11. 17284
    Prev
    Next
    INDEX
    Explanations

    phrases with a sarcastic or mocking tone

    oai_token-act-pair · gpt-3.5-turbo

    instances of the text "sn" which appears to relate to themes of sarcasm or dark humor

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

    No Comments

    Negative Logits
    heid
    -0.88
    xual
    -0.73
    mine
    -0.72
    EMENT
    -0.71
    shire
    -0.66
     PowerPoint
    -0.65
    WAYS
    -0.61
    minus
    -0.61
    Constructed
    -0.60
    geist
    -0.59
    POSITIVE LOGITS
    agging
    1.14
    atching
    1.14
    agged
    1.12
    obb
    1.11
    atches
    1.07
    ugg
    1.05
    appy
    1.03
    apper
    1.02
    appers
    1.00
    apping
    0.98
    Activations Density 0.008%

    No Known Activations