© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 17650
    Prev
    Next
    INDEX
    Explanations

    mentions and details about former officials and their roles in security-related positions

    oai_token-act-pair · gpt-3.5-turbo

    references to former government officials and their roles

    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
     masc
    -0.91
     eviction
    -0.79
    unicip
    -0.79
    xual
    -0.79
     crowdfunding
    -0.77
    enfranch
    -0.77
    bucks
    -0.76
     storefront
    -0.73
     corrid
    -0.73
     infringing
    -0.71
    POSITIVE LOGITS
     declass
    1.54
     Petraeus
    1.52
     Kissinger
    1.50
     Pentagon
    1.49
     CIA
    1.45
    CIA
    1.45
     Clapper
    1.44
     neocons
    1.42
     Mattis
    1.36
     Cheney
    1.29
    Activations Density 0.681%

    No Known Activations