© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 14678
    Prev
    Next
    INDEX
    Explanations

    references to the greenhouse gas carbon dioxide

    oai_token-act-pair · gpt-3.5-turbo

    references to carbon dioxide levels

    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
     Styles
    -0.66
    her
    -0.66
     Br
    -0.64
     Fort
    -0.63
     Aff
    -0.62
    br
    -0.60
     Ranked
    -0.60
     Barber
    -0.60
    friend
    -0.59
     Mend
    -0.59
    POSITIVE LOGITS
     dioxide
    3.49
    ioxide
    2.34
     oxide
    1.59
     gases
    1.14
     pollut
    1.13
     chloride
    1.10
     pollutants
    1.05
    monary
    1.05
     methane
    1.05
    elsius
    1.03
    Activations Density 0.013%

    No Known Activations