© 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 768
    9. 8-RES_FS768-JB
    11. 725
    Prev
    Next
    INDEX
    Explanations

    words related to surveys and polls

    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
    490
    +0.11
    0.5%
    167
    +0.10
    0.5%
    651
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    540
    +0.11
    0.17
    288
    +0.10
    0.19
    651
    +0.10
    0.14
    Negative Logits
     Leilan
    -0.88
    ModLoader
    -0.82
    ulkan
    -0.69
    osi
    -0.68
    anti
    -0.67
    joy
    -0.66
    illard
    -0.66
    adia
    -0.65
    AFTA
    -0.64
     goodbye
    -0.64
    POSITIVE LOGITS
     methodology
    0.91
     data
    0.90
     results
    0.87
    results
    0.86
    itatively
    0.85
     evaluations
    0.85
    iologist
    0.84
    iological
    0.83
     assessing
    0.83
     report
    0.82
    Activations Density 5.357%

    No Known Activations