© 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. 22115
    Prev
    Next
    INDEX
    Explanations

    conditional phrases and modal verbs indicating obligation or potential actions

    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
    551
    +0.12
    0.5%
    540
    +0.09
    0.4%
    447
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.12
    0.07
    566
    +0.09
    0.05
    86
    +0.09
    0.05
    Negative Logits
     Niet
    -0.81
     intrig
    -0.78
    mess
    -0.77
     mir
    -0.75
     theories
    -0.71
     coincidence
    -0.68
    soType
    -0.68
    wow
    -0.67
    mosp
    -0.66
     myths
    -0.66
    POSITIVE LOGITS
     ineligible
    1.13
     exempt
    1.09
     exempted
    1.06
     automatically
    1.04
     eligible
    1.01
     concess
    0.99
     barred
    0.96
     forfeit
    0.96
     prohibited
    0.91
     revert
    0.91
    Activations Density 0.333%

    No Known Activations