© 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. 0-RES-JB
    11. 5248
    Prev
    Next
    INDEX
    Explanations

    words related to significant negative impacts or events

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "devastating" and related variations signifying significant damage or distress

    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
    uty
    -0.66
    stad
    -0.66
    sburg
    -0.66
     /
    -0.66
     Attend
    -0.64
     Orient
    -0.63
     ®
    -0.62
     Anyway
    -0.62
    Desk
    -0.62
    '/
    -0.61
    POSITIVE LOGITS
     devastating
    3.35
     crippling
    2.03
     disastrous
    1.99
     heartbreaking
    1.98
     debilitating
    1.94
     devast
    1.93
     catastrophic
    1.87
     devastation
    1.83
     horrific
    1.78
     devastated
    1.77
    Activations Density 0.020%

    No Known Activations