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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & 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. 277
    Prev
    Next
    INDEX
    Explanations

    phrases related to expressing personal opinions or perspectives

    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
    107
    +0.10
    0.5%
    411
    +0.10
    0.5%
    702
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.10
    0.17
    243
    +0.10
    0.15
    715
    +0.10
    0.13
    Negative Logits
     themselves
    -0.69
     Hebdo
    -0.64
     dissip
    -0.64
     allegedly
    -0.62
     dissolved
    -0.61
     purportedly
    -0.60
     itself
    -0.60
     ousted
    -0.60
     Us
    -0.59
    emer
    -0.59
    POSITIVE LOGITS
     myself
    1.68
     personally
    1.04
    poke
    0.98
     my
    0.94
     glad
    0.89
     regret
    0.82
     thankful
    0.80
     fortunate
    0.79
     recommend
    0.78
     wondering
    0.76
    Activations Density 6.077%

    No Known Activations