© 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. 8-RES-JB
    11. 2489
    Prev
    Next
    INDEX
    Explanations

    words related to international politics and sanctions against countries

    oai_token-act-pair · gpt-3.5-turbo

    mentions of military actions and nuclear proliferation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Died
    -0.94
     Benson
    -0.80
    aucus
    -0.77
     "$:/
    -0.76
     Blackwell
    -0.75
     Bronx
    -0.74
     Compton
    -0.73
    olesterol
    -0.73
     Blasio
    -0.70
     Philly
    -0.69
    POSITIVE LOGITS
     aggression
    1.05
     nuclear
    1.04
     territorial
    1.03
     Lavrov
    1.03
     destabil
    1.03
     retaliate
    1.01
     retali
    1.00
     annexed
    0.99
     missile
    0.97
     sovereignty
    0.96
    Activations Density 0.568%

    No Known Activations