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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 371
    Prev
    Next
    INDEX
    Explanations

    information related to renewable energy, military interest in solar power, public appreciation for assistance, and charitable contributions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    481
    +0.73
    7.4%
    447
    +0.34
    3.4%
    393
    +0.07
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    481
    +0.73
    0.33
    393
    +0.34
    0.27
    703
    +0.07
    0.05
    Negative Logits
    ãĤ§
    -0.33
    anity
    -0.33
    ogenesis
    -0.30
    hya
    -0.30
    osc
    -0.28
    ulia
    -0.28
    ilib
    -0.28
     outweigh
    -0.28
    esame
    -0.28
    aly
    -0.28
    POSITIVE LOGITS
     similarly
    0.28
     direct
    0.28
     Selected
    0.28
     Saw
    0.26
     referen
    0.25
     upper
    0.24
     Allied
    0.24
     recall
    0.24
     HK
    0.24
     Quote
    0.24
    Activations Density 11.399%

    No Known Activations