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 49152
    5. 8-RES_FS49152-JB
    6. 26033
    Prev
    Next
    INDEX
    Explanations

    terms related to the creation and establishment of organizations or initiatives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    723
    +0.11
    0.5%
    766
    +0.10
    0.5%
    454
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    59
    +0.11
    0.03
    766
    +0.10
    0.03
    529
    +0.10
    0.03
    Negative Logits
    âĨij
    -0.84
    eches
    -0.73
     impacting
    -0.69
    mol
    -0.68
    WARE
    -0.67
     bites
    -0.66
    iev
    -0.65
    itta
    -0.65
    ibble
    -0.64
    loading
    -0.63
    POSITIVE LOGITS
     committees
    0.81
     Patreon
    0.79
     memorial
    0.76
     nonpartisan
    0.76
     committee
    0.74
     dedicated
    0.74
     Task
    0.73
     bipartisan
    0.73
     parallel
    0.72
     safeguards
    0.72
    Activations Density 0.086%

    No Known Activations