© 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. 2-RES-JB
    11. 21457
    Prev
    Next
    INDEX
    Explanations

    references to geographical directions, specifically the north

    oai_token-act-pair · gpt-3.5-turbo

    references to the direction "north."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    TABLE
    -0.81
    inventoryQuantity
    -0.81
     CLSID
    -0.80
    PRESS
    -0.79
    TYPE
    -0.78
    wcsstore
    -0.74
    ATURE
    -0.73
    tons
    -0.72
    aminer
    -0.72
    Definition
    -0.72
    POSITIVE LOGITS
    ward
    1.21
    west
    0.99
    side
    0.98
    shore
    0.97
     west
    0.96
     east
    0.95
     hemisphere
    0.93
    western
    0.92
    wards
    0.90
    bound
    0.87
    Activations Density 0.016%

    No Known Activations