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. 4-RES-JB
    6. 9839
    Prev
    Next
    INDEX
    Explanations

    locations or landmarks, specifically focusing on Downing Street and Buckingham Palace

    oai_token-act-pair · gpt-3.5-turbo

    mention of specific locations or landmarks

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

    No Comments

    Negative Logits
    Magikarp
    -0.76
    ually
    -0.75
    Reviewer
    -0.73
    ersion
    -0.71
    uality
    -0.71
     Gutenberg
    -0.71
    ENDED
    -0.70
    hammad
    -0.69
    bnb
    -0.68
    Discover
    -0.67
    POSITIVE LOGITS
     Downing
    1.07
    nuts
    0.97
    nut
    0.87
    minster
    0.84
    hurst
    0.80
    tree
    0.78
     Abbey
    0.77
    vous
    0.72
    flower
    0.71
     MX
    0.71
    Activations Density 0.007%

    No Known Activations