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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 565
    Prev
    Next
    INDEX
    Explanations

    instances where someone was dismissed from a position or faced legal charges

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    566
    +0.14
    0.7%
    102
    +0.14
    0.7%
    297
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.14
    0.20
    102
    +0.14
    0.19
    566
    +0.13
    0.16
    Negative Logits
    inav
    -0.73
    Footnote
    -0.72
     Extend
    -0.69
    iates
    -0.69
     Matters
    -0.68
     Which
    -0.66
     IMAGES
    -0.66
     HAVE
    -0.66
     Compare
    -0.65
     Millions
    -0.65
    POSITIVE LOGITS
    hes
    1.19
     able
    1.17
     supposed
    1.16
    wolves
    1.13
     born
    1.12
     originally
    1.05
     unable
    1.03
    nt
    1.00
    wolf
    1.00
     instrumental
    0.98
    Activations Density 1.353%

    No Known Activations