© 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. 9763
    Prev
    Next
    INDEX
    Explanations

    phrases related to political and economic systems and issues

    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
    219
    +0.11
    0.5%
    540
    +0.09
    0.4%
    418
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.11
    0.05
    278
    +0.09
    0.04
    335
    +0.09
    0.04
    Negative Logits
    etsk
    -0.61
    oping
    -0.60
    ograms
    -0.59
    ikan
    -0.58
    oop
    -0.58
    axter
    -0.58
    aunch
    -0.58
    avery
    -0.57
    alks
    -0.57
     Shooting
    -0.57
    POSITIVE LOGITS
     motiv
    1.14
     motivating
    1.13
     factor
    1.09
     hurdle
    1.08
     impetus
    1.07
     deterrent
    1.06
     prerequisite
    1.05
     concern
    1.00
     motivation
    0.97
     indicator
    0.96
    Activations Density 0.273%

    No Known Activations