© 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. 5265
    Prev
    Next
    INDEX
    Explanations

    the phrase "long been" indicating an extended duration or historical context

    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
    532
    +0.10
    0.4%
    373
    +0.09
    0.4%
    130
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    344
    +0.10
    0.04
    463
    +0.09
    0.03
    423
    +0.09
    0.03
    Negative Logits
    umbn
    -0.69
    toc
    -0.68
     attm
    -0.67
    natureconservancy
    -0.63
    lations
    -0.61
    xit
    -0.60
     dstg
    -0.60
    ithering
    -0.58
    UGH
    -0.58
    rop
    -0.57
    POSITIVE LOGITS
     dreamed
    0.96
     regarded
    0.94
     wondered
    0.83
     doubted
    0.80
     touted
    0.78
     believed
    0.77
     considered
    0.76
     admired
    0.75
     feared
    0.75
     tradition
    0.74
    Activations Density 0.083%

    No Known Activations