© 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. 22127
    Prev
    Next
    INDEX
    Explanations

    the continuous tense form of verbs, specifically those ending in 'ing'

    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
    480
    +0.15
    0.7%
    571
    +0.13
    0.6%
    154
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    571
    +0.15
    0.01
    464
    +0.13
    0.01
    524
    +0.10
    0.01
    Negative Logits
     Seat
    -0.89
    ãĥ¯ãĥ³
    -0.70
    è¦ļéĨĴ
    -0.69
     BMC
    -0.67
     Gamble
    -0.65
     Pryor
    -0.64
     Companion
    -0.64
    女
    -0.62
     Jagu
    -0.61
     Recall
    -0.60
    POSITIVE LOGITS
    redients
    1.44
    raved
    0.97
    ames
    0.94
    lass
    0.90
    ame
    0.89
    ertain
    0.85
    oing
    0.84
    lasses
    0.83
    ulated
    0.82
    ert
    0.81
    Activations Density 0.022%

    No Known Activations