© 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. 5260
    Prev
    Next
    INDEX
    Explanations

    numerical values, particularly in a structured format

    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
    314
    +0.10
    0.5%
    71
    +0.10
    0.5%
    87
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.10
    0.03
    565
    +0.10
    0.03
    316
    +0.10
    0.03
    Negative Logits
     numbering
    -0.85
     mathemat
    -0.82
     neighb
    -0.79
    undai
    -0.76
     contrace
    -0.76
     unden
    -0.72
     totality
    -0.72
     apprentices
    -0.72
     ensu
    -0.70
     disadvant
    -0.70
    POSITIVE LOGITS
    888
    0.97
    269
    0.94
    328
    0.93
    687
    0.92
    198
    0.92
    875
    0.91
    008
    0.91
    268
    0.91
    016
    0.90
    378
    0.90
    Activations Density 0.073%

    No Known Activations