© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 19466
    Prev
    Next
    INDEX
    Explanations

    words related to political figures, plans for the future, and official statements

    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
    665
    +0.17
    0.8%
    447
    +0.11
    0.5%
    354
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.17
    0.06
    49
    +0.11
    0.04
    339
    +0.10
    0.03
    Negative Logits
    quickShipAvailable
    -0.76
    worthy
    -0.68
     Explan
    -0.68
    ritical
    -0.68
    clips
    -0.61
    pite
    -0.60
    checking
    -0.59
     sensibilities
    -0.58
     Canary
    -0.58
     comprehension
    -0.58
    POSITIVE LOGITS
     pursue
    1.14
     revise
    1.13
     introduce
    1.09
     utilize
    1.09
     propose
    1.08
     continue
    1.07
     publish
    1.07
     invest
    1.07
     donate
    1.05
     retire
    1.05
    Activations Density 0.096%

    No Known Activations