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 768
    5. 8-RES_FS768-JB
    6. 103
    Prev
    Next
    INDEX
    Explanations

    keywords related to entrepreneurship and business activities

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    +0.17
    0.8%
    330
    +0.13
    0.6%
    575
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.17
    0.18
    330
    +0.13
    0.17
    537
    +0.12
    0.17
    Negative Logits
    ASED
    -0.64
    FLAG
    -0.61
    Prosecut
    -0.59
    rawdownloadcloneembedreportprint
    -0.59
    OLOG
    -0.59
    yss
    -0.59
    CHECK
    -0.58
     bilateral
    -0.58
     Territories
    -0.57
    OLOGY
    -0.57
    POSITIVE LOGITS
    hips
    1.70
    hip
    1.70
    paces
    1.47
     who
    1.27
    '
    1.22
    pace
    1.22
    ervatives
    1.20
    folk
    1.19
    mith
    1.10
     alike
    1.10
    Activations Density 3.252%

    No Known Activations