© 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. 267
    Prev
    Next
    INDEX
    Explanations

    chemicals and substances that may be harmful to health

    oai_token-act-pair · gpt-3.5-turbo

    references to hazardous chemicals and their effects

    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
    ership
    -0.82
    thumbnails
    -0.78
    ļé
    -0.69
    ģĸ
    -0.68
     circumference
    -0.64
    iversary
    -0.61
    ¯
    -0.61
    ĺħ
    -0.61
    cember
    -0.61
    square
    -0.61
    POSITIVE LOGITS
    ervatives
    1.03
     emitted
    1.01
     ingested
    0.99
     extracted
    0.99
    ervative
    0.98
     sprayed
    0.94
     additives
    0.94
     such
    0.92
     contamin
    0.91
     chemically
    0.90
    Activations Density 0.203%

    No Known Activations