© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4902
    Prev
    Next
    INDEX
    Explanations

    references to bowls and bowl games

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    128
    +0.12
    0.6%
    484
    +0.09
    0.4%
    377
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.12
    0.02
    288
    +0.09
    0.02
    40
    +0.09
    0.01
    Negative Logits
    kson
    -0.88
     Lauder
    -0.76
     Jenner
    -0.64
     Bundes
    -0.64
     Hugo
    -0.64
     Draper
    -0.64
    bourg
    -0.62
    eanor
    -0.61
     Assass
    -0.60
     Benson
    -0.60
    POSITIVE LOGITS
    pipe
    1.00
    ers
    0.95
    cup
    0.92
    strings
    0.88
    bowl
    0.87
    hend
    0.85
    plot
    0.81
    chair
    0.81
    gary
    0.80
    mates
    0.80
    Activations Density 0.052%

    No Known Activations