© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 1536
    9. 8-RES_FS1536-JB
    11. 5
    Prev
    Next
    INDEX
    Explanations

    positive evaluations or value judgments

    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
    51
    +0.12
    0.6%
    162
    +0.11
    0.5%
    128
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    51
    +0.12
    0.10
    162
    +0.11
    0.09
    628
    +0.09
    0.08
    Negative Logits
    eters
    -0.78
    ategory
    -0.75
    hyde
    -0.71
    ptin
    -0.70
    oths
    -0.66
    eds
    -0.64
    opers
    -0.64
    gemony
    -0.63
     Tsukuyomi
    -0.62
    atum
    -0.61
    POSITIVE LOGITS
     luck
    1.15
    enough
    1.13
    bye
    1.11
    luck
    1.06
     intentions
    1.00
     Samar
    0.98
     ol
    0.94
    reads
    0.93
     manners
    0.89
    nat
    0.89
    Activations Density 1.529%

    No Known Activations