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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 6147
    Prev
    Next
    INDEX
    Explanations

    possessive forms and references to organizations or groups

    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
    143
    +0.10
    0.5%
    594
    +0.10
    0.4%
    184
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.10
    0.11
    608
    +0.10
    0.10
    616
    +0.10
    0.07
    Negative Logits
    android
    -0.77
    cedented
    -0.72
    gdala
    -0.72
    bucks
    -0.71
    $$$$
    -0.71
    tch
    -0.69
    cox
    -0.68
    zon
    -0.68
    anges
    -0.67
    riots
    -0.65
    POSITIVE LOGITS
     founder
    1.34
     founders
    1.30
     founding
    1.18
     CEO
    1.02
     logo
    1.01
     headquarters
    1.00
     chairman
    0.99
     owner
    0.98
     leader
    0.97
     directors
    0.97
    Activations Density 1.276%

    No Known Activations