Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 15266
    Prev
    Next
    INDEX
    Explanations

    words related to government intelligence agencies and whistleblowers

    oai_token-act-pair · gpt-3.5-turbo

    references to Edward Snowden and related events or details about him

    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
    bowl
    -0.84
    entary
    -0.84
     offic
    -0.76
     Veterinary
    -0.75
     seating
    -0.75
     counties
    -0.74
     Towns
    -0.73
    ã
    -0.73
     Hort
    -0.72
     bowl
    -0.72
    POSITIVE LOGITS
     WikiLeaks
    1.28
     Wikileaks
    1.25
    ileaks
    1.17
     whistlebl
    1.13
     trove
    1.13
    Leaks
    1.12
     revelations
    1.11
     hacked
    1.10
     Snowden
    1.09
     disclosures
    1.08
    Activations Density 0.540%

    No Known Activations