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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 3072
    9. 8-RES_FS3072-JB
    11. 199
    Prev
    Next
    INDEX
    Explanations

    past tense verbs as well as phrases related to events or actions happening

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    481
    +0.14
    0.7%
    408
    +0.10
    0.5%
    757
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    172
    +0.14
    0.09
    375
    +0.10
    0.06
    59
    +0.10
    0.07
    Negative Logits
    ãĤ´ãĥ³
    -0.78
    bda
    -0.76
    cffffcc
    -0.70
    ãĤ§
    -0.70
    inav
    -0.69
    ãģį
    -0.65
    displayText
    -0.64
    express
    -0.62
    alone
    -0.61
    represent
    -0.61
    POSITIVE LOGITS
     someone
    1.02
     people
    0.95
     somebody
    0.92
     landlords
    0.91
     individuals
    0.87
     technicians
    0.86
     governments
    0.86
     anyone
    0.86
     users
    0.86
     they
    0.82
    Activations Density 2.222%

    No Known Activations