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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 2326
    Prev
    Next
    INDEX
    Explanations

    information related to financial advice and investment strategies, particularly focusing on Roth IRAs

    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
    640
    +0.38
    2.0%
    160
    +0.17
    0.9%
    107
    +0.15
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.38
    0.26
    107
    +0.17
    0.22
    160
    +0.15
    0.24
    Negative Logits
     sculpt
    -0.81
     neighb
    -0.79
     spitting
    -0.77
     casc
    -0.77
     frontline
    -0.75
     thrill
    -0.74
     swe
    -0.73
     marching
    -0.73
     sucker
    -0.72
     spir
    -0.72
    POSITIVE LOGITS
    Conclusion
    1.78
    Additionally
    1.64
    Advertisements
    1.62
    Also
    1.59
    References
    1.59
    Furthermore
    1.58
    CONCLUS
    1.57
    Lastly
    1.55
    Other
    1.54
    However
    1.54
    Activations Density 1.193%

    No Known Activations