Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 184
    Prev
    Next
    INDEX
    Explanations

    keywords related to climate change, environmental impact, pollution, and sustainable energy sources

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    644
    +0.11
    0.5%
    12
    +0.10
    0.5%
    284
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.14
    57
    +0.10
    0.06
    575
    +0.10
    0.09
    Negative Logits
    dom
    -0.84
    doms
    -0.80
    loading
    -0.77
    lique
    -0.77
    staff
    -0.69
    minster
    -0.69
    amina
    -0.69
    ime
    -0.69
    lisher
    -0.69
    IME
    -0.68
    POSITIVE LOGITS
     dioxide
    1.05
     hazards
    0.98
     pollution
    0.95
     warming
    0.95
     climate
    0.95
     impacts
    0.91
     Climate
    0.90
     disasters
    0.89
     footprint
    0.87
     sustainability
    0.87
    Activations Density 2.185%

    No Known Activations