© 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. 23883
    Prev
    Next
    INDEX
    Explanations

    references to progress and achievements in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered 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
    447
    +0.14
    0.6%
    481
    +0.13
    0.6%
    349
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    724
    +0.14
    0.05
    16
    +0.13
    0.03
    349
    +0.09
    0.03
    Negative Logits
    pron
    -0.77
    umed
    -0.75
    umes
    -0.74
    anity
    -0.69
     denotes
    -0.68
    $$$$
    -0.68
     airs
    -0.67
    nis
    -0.67
     spoilers
    -0.67
     vomit
    -0.67
    POSITIVE LOGITS
     achievements
    1.33
     accomplishments
    1.25
     successes
    1.23
     strides
    1.17
     achievement
    1.15
     accomplishment
    1.10
     advancements
    1.07
     progress
    1.07
     achieved
    1.06
     improvements
    1.04
    Activations Density 0.573%

    No Known Activations