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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9-RES-JB
    11. 2576
    Prev
    Next
    INDEX
    Explanations

    locations, especially government buildings and landmarks

    oai_token-act-pair · gpt-3.5-turbo

    references to significant political or governmental locations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    adapt
    -0.72
    CAST
    -0.72
    lass
    -0.68
    lust
    -0.68
    apter
    -0.68
    dit
    -0.66
    nir
    -0.65
    rounder
    -0.65
    GM
    -0.65
    loss
    -0.65
    POSITIVE LOGITS
     entrances
    1.08
     plaza
    1.03
     Visitors
    1.01
     Plaza
    1.01
     entrance
    1.00
     balcony
    0.99
     premises
    0.98
     chapel
    0.96
     Hotel
    0.95
     Building
    0.94
    Activations Density 0.199%

    No Known Activations