© 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 24576
    9. 8-RES_FS24576-JB
    11. 9269
    Prev
    Next
    INDEX
    Explanations

    references to specific measurements or numerical values, such as years or distances

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    127
    +0.09
    0.4%
    6
    +0.09
    0.4%
    511
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    260
    +0.09
    0.05
    462
    +0.09
    0.04
    639
    +0.09
    0.04
    Negative Logits
    aida
    -0.72
    sters
    -0.67
    inki
    -0.67
    vae
    -0.66
     Kodi
    -0.66
    anski
    -0.65
    repre
    -0.65
     bait
    -0.64
    ndra
    -0.63
     blu
    -0.61
    POSITIVE LOGITS
    degree
    0.95
    ãĤ¼ãĤ¦ãĤ¹
    0.91
    percent
    0.88
     MPH
    0.88
    168
    0.84
    ãĥīãĥ©ãĤ´ãĥ³
    0.82
    230
    0.82
    616
    0.81
    士
    0.81
     Percent
    0.80
    Activations Density 0.135%

    No Known Activations