© 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. 8050
    Prev
    Next
    INDEX
    Explanations

    mentions of political figures and leaders

    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
    447
    +0.12
    0.5%
    104
    +0.11
    0.5%
    735
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.12
    0.06
    247
    +0.11
    0.04
    291
    +0.11
    0.04
    Negative Logits
    76561
    -0.69
    ighters
    -0.67
    utenberg
    -0.64
    aves
    -0.61
    items
    -0.58
     contents
    -0.57
     Consumers
    -0.57
    Items
    -0.57
     bins
    -0.56
     Adren
    -0.56
    POSITIVE LOGITS
     whom
    1.06
     who
    1.00
     named
    0.97
     willing
    0.93
     whose
    0.85
     persona
    0.85
    who
    0.84
     capable
    0.83
     overseeing
    0.83
    hunt
    0.79
    Activations Density 0.307%

    No Known Activations