© 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 49152
    9. 8-RES_FS49152-JB
    11. 41592
    Prev
    Next
    INDEX
    Explanations

    words related to regulatory and oversight organizations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    629
    +0.10
    0.5%
    419
    +0.10
    0.5%
    146
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.10
    0.04
    401
    +0.10
    0.03
    419
    +0.10
    0.03
    Negative Logits
    brew
    -0.81
    Ô
    -0.79
    å¦
    -0.69
    kell
    -0.68
    ãĤª
    -0.68
    ãĥĥãĤ¯
    -0.67
    erenn
    -0.67
    çĶŁ
    -0.66
    Guest
    -0.66
    ixel
    -0.65
    POSITIVE LOGITS
     IG
    0.84
     overseeing
    0.75
     tasked
    0.74
     SOC
    0.73
     Edu
    0.73
     FAT
    0.72
    arians
    0.72
     Carmen
    0.72
     Deutsche
    0.70
     Ros
    0.70
    Activations Density 0.075%

    No Known Activations