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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlog/PodcastSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 3866
    Prev
    Next
    INDEX
    Explanations

    words related to legal or official ceremonies, specifically involving the act of being sworn in

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "sworn" and its variations, indicating references to oaths or formal affirmations

    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
     univers
    -0.70
    Boost
    -0.70
    Stud
    -0.68
    tainment
    -0.67
    activity
    -0.66
     blocks
    -0.65
    Games
    -0.65
    ost
    -0.64
     Blocks
    -0.63
     overload
    -0.63
    POSITIVE LOGITS
     sworn
    4.35
     swore
    1.93
     swearing
    1.70
     oath
    1.65
     swear
    1.63
     Oath
    1.37
     affidav
    1.23
     affidavit
    1.08
     elected
    1.04
    elected
    0.99
    Activations Density 0.005%

    No Known Activations