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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 8-RES-JB
    11. 6347
    Prev
    Next
    INDEX
    Explanations

    references to specific locations, particularly New York City

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "New York City."

    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
    rar
    -0.91
    iru
    -0.79
    ãĥł
    -0.78
    VALUE
    -0.76
    Wan
    -0.76
    insk
    -0.75
    ãĥĦ
    -0.74
    20439
    -0.74
    ãĥ¼ãĥĨ
    -0.74
    iago
    -0.73
    POSITIVE LOGITS
     landmarks
    0.96
     borough
    0.94
     subway
    0.90
     FC
    0.88
     streets
    0.85
     Orchestra
    0.83
     skyline
    0.82
     neighborhoods
    0.82
     Opera
    0.82
     skysc
    0.82
    Activations Density 0.097%

    No Known Activations