© 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. 10-RES_POST_32K-OAI
    11. 30417
    Prev
    Next
    INDEX
    Explanations

    financial terms or figures related to healthcare spending

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    102
    +0.12
    0.5%
    196
    +0.12
    0.5%
    260
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    160
    +0.12
    0.17
    102
    +0.12
    0.15
    107
    +0.11
    0.13
    Negative Logits
     premise
    -0.68
     indo
    -0.68
     trope
    -0.64
     vice
    -0.64
     psyche
    -0.62
     profession
    -0.61
    malink
    -0.61
     fundament
    -0.61
     END
    -0.58
     volunt
    -0.58
    POSITIVE LOGITS
    approximately
    1.61
    300
    1.53
    eight
    1.48
    600
    1.45
    900
    1.44
    150
    1.43
    250
    1.43
    400
    1.42
    750
    1.42
    500
    1.40
    Activations Density 0.633%

    No Known Activations