© 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. 5020
    Prev
    Next
    INDEX
    Explanations

    mentions of the Cold War

    oai_token-act-pair · gpt-3.5-turbo

    references to the "Cold War."

    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
    merce
    -0.81
     tou
    -0.71
    à¼
    -0.71
     gou
    -0.66
    orial
    -0.66
    yrus
    -0.65
     nib
    -0.64
    phabet
    -0.64
     enrol
    -0.62
     succeed
    -0.62
    POSITIVE LOGITS
     Cold
    3.97
    Cold
    3.14
    cold
    2.43
     cold
    2.03
     colder
    1.35
     chilly
    1.32
     Warm
    1.22
     Cool
    1.19
     Hot
    1.18
     Chill
    1.16
    Activations Density 0.015%

    No Known Activations