© 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. 5500
    Prev
    Next
    INDEX
    Explanations

    sequences indicating financial transactions or economic activities

    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
    +0.14
    0.6%
    373
    +0.10
    0.5%
    309
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.14
    0.00
    284
    +0.10
    0.02
    530
    +0.09
    0.01
    Negative Logits
    ŃĶ
    -0.64
     surplus
    -0.64
    ²¾
    -0.58
    oj
    -0.58
     dissatisf
    -0.57
    econom
    -0.57
     corros
    -0.54
     bottleneck
    -0.54
     tidal
    -0.52
     workforce
    -0.52
    POSITIVE LOGITS
    ummies
    1.29
     sake
    1.12
    gotten
    1.11
     purposes
    0.91
     reasons
    0.81
    earance
    0.80
    ktop
    0.79
     hire
    0.77
    arcer
    0.76
     ages
    0.74
    Activations Density 0.128%

    No Known Activations