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. 917
    Prev
    Next
    INDEX
    Explanations

    economic and financial terms and information

    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
    668
    +0.10
    0.4%
    552
    +0.09
    0.4%
    330
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    66
    +0.10
    0.10
    570
    +0.09
    0.00
    471
    +0.09
    0.07
    Negative Logits
     oath
    -0.86
     indoctr
    -0.83
     commandments
    -0.73
     befriend
    -0.73
     indo
    -0.73
     guilt
    -0.73
     spies
    -0.72
     torture
    -0.72
     traitor
    -0.71
     voic
    -0.71
    POSITIVE LOGITS
    flation
    1.13
     forecast
    1.08
     forecasts
    1.08
     earnings
    1.05
    Chart
    1.04
     Adjusted
    1.03
     GDP
    1.01
     inflation
    1.01
    adjusted
    1.01
     depreciation
    0.98
    Activations Density 4.538%

    No Known Activations