© 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. 7309
    Prev
    Next
    INDEX
    Explanations

    technical terms and financial information

    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
    443
    +0.15
    0.7%
    67
    +0.12
    0.6%
    373
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    427
    +0.15
    0.02
    443
    +0.12
    0.02
    540
    +0.11
    0.03
    Negative Logits
     ultras
    -0.88
    ateurs
    -0.84
    etus
    -0.74
     Gaia
    -0.72
    isine
    -0.72
    ĸļ
    -0.71
     controvers
    -0.71
    dding
    -0.70
    toggle
    -0.70
    oldown
    -0.69
    POSITIVE LOGITS
     herein
    1.01
     Company
    0.99
     Compensation
    0.94
     subsidiaries
    0.92
     derivative
    0.91
     hereby
    0.90
     Agreement
    0.90
     Securities
    0.89
     goodwill
    0.87
     Investor
    0.84
    Activations Density 0.178%

    No Known Activations