© 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. 10-RES-JB
    11. 17517
    Prev
    Next
    INDEX
    Explanations

    mentions of countries and their nuclear programs, specifically focused on Iran and North Korea

    oai_token-act-pair · gpt-3.5-turbo

    references to nuclear weapons and associated geopolitical concerns

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Blasio
    -0.93
    aucus
    -0.86
    soDeliveryDate
    -0.82
    visors
    -0.81
    soType
    -0.78
     ILCS
    -0.75
    HAEL
    -0.75
     Franch
    -0.75
     Volunte
    -0.74
     Clarence
    -0.73
    POSITIVE LOGITS
     nuclear
    1.27
    nuclear
    1.13
     missiles
    1.10
     atomic
    1.09
     missile
    1.09
     retaliate
    1.03
     stockpile
    1.03
     Nuclear
    1.00
     threaten
    0.99
     annex
    0.99
    Activations Density 0.358%

    No Known Activations