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. 11-RES-JB
    6. 668
    Prev
    Next
    INDEX
    Explanations

    instances of sentences talking about legal or official actions involving specific individuals

    oai_token-act-pair · gpt-3.5-turbo

    references to actions or decisions made by authorities or governing bodies

    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
    verage
    -0.72
    uality
    -0.71
    eware
    -0.70
    ppo
    -0.69
    bits
    -0.69
    LP
    -0.69
    inea
    -0.68
    apy
    -0.68
    aunder
    -0.65
    empl
    -0.65
    POSITIVE LOGITS
    laws
    1.16
     virtue
    1.10
     decree
    1.09
     President
    1.02
     superiors
    1.00
     Parliament
    0.99
     parliament
    0.98
     Congress
    0.96
     Governor
    0.95
     statute
    0.94
    Activations Density 0.110%

    No Known Activations