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. 660
    Prev
    Next
    INDEX
    Explanations

    phrases related to scientific research and organisms

    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
    447
    +0.14
    0.6%
    496
    +0.12
    0.5%
    642
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    469
    +0.14
    0.10
    665
    +0.12
    0.13
    677
    +0.10
    0.10
    Negative Logits
    aga
    -0.59
    Ī
    -0.57
    IJ
    -0.56
    eries
    -0.56
    Ĺ
    -0.54
    wings
    -0.54
    ritical
    -0.53
    tan
    -0.53
    brid
    -0.52
    tons
    -0.51
    POSITIVE LOGITS
     elsewhere
    1.01
     anywhere
    1.00
     alongside
    0.97
     everywhere
    0.92
     outdoors
    0.91
     concurrently
    0.89
     throughout
    0.88
     atop
    0.86
     here
    0.85
     across
    0.84
    Activations Density 2.844%

    No Known Activations