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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 1369
    Prev
    Next
    INDEX
    Explanations

    references to legal charges and trials

    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
    288
    +0.10
    0.4%
    447
    +0.09
    0.4%
    268
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    386
    +0.10
    0.10
    288
    +0.09
    0.10
    540
    +0.09
    0.09
    Negative Logits
    birth
    -0.79
    ARCH
    -0.74
    toggle
    -0.72
    orph
    -0.70
    skill
    -0.69
    reen
    -0.69
    patch
    -0.66
    population
    -0.65
    atomic
    -0.65
    psey
    -0.65
    POSITIVE LOGITS
    heet
    1.19
     accusing
    0.95
     wrongdoing
    0.93
     alleging
    0.88
     accused
    0.86
     accuses
    0.83
    llah
    0.83
     accuse
    0.80
     levied
    0.78
     Charges
    0.78
    Activations Density 1.572%

    No Known Activations