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. 0-RES-JB
    6. 14482
    Prev
    Next
    INDEX
    Explanations

    phrases related to a specific location - Houston, Texas

    oai_token-act-pair · gpt-3.5-turbo

    references to the city of Houston and its associated sports teams

    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
     monks
    -0.78
     Nato
    -0.77
    Vo
    -0.77
     monk
    -0.75
    Versions
    -0.74
     np
    -0.74
     inval
    -0.72
     goblins
    -0.70
     armour
    -0.69
    vati
    -0.69
    POSITIVE LOGITS
     Houston
    3.63
    Houston
    3.30
     Texans
    2.05
     Dallas
    2.01
     Texas
    1.93
     Astros
    1.87
    Dallas
    1.86
     HOU
    1.85
    Texas
    1.83
    ouston
    1.82
    Activations Density 0.020%

    No Known Activations