© 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. 4889
    Prev
    Next
    INDEX
    Explanations

    instances of the word "subject" followed by different implications or conditions

    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
    352
    +0.09
    0.4%
    643
    +0.09
    0.4%
    155
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    609
    +0.09
    0.02
    541
    +0.09
    0.02
    118
    +0.09
    0.01
    Negative Logits
     ect
    -0.62
    yip
    -0.60
     sake
    -0.59
     slit
    -0.57
    cia
    -0.56
     ii
    -0.55
     pitch
    -0.55
     passer
    -0.54
    roma
    -0.54
     (>
    -0.52
    POSITIVE LOGITS
    ioned
    0.75
    quartered
    0.75
    achusetts
    0.74
    ivating
    0.73
    externalActionCode
    0.69
    lined
    0.67
    ively
    0.66
    uted
    0.64
    ãĥŀ
    0.62
    ivated
    0.62
    Activations Density 0.077%

    No Known Activations