© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 26597
    Prev
    Next
    INDEX
    Explanations

    numerical data related to financial performance or market trends

    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
    380
    +0.11
    0.5%
    684
    +0.10
    0.5%
    577
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    577
    +0.11
    0.01
    407
    +0.10
    0.01
    534
    +0.10
    0.01
    Negative Logits
    orno
    -0.63
    ????????
    -0.61
    Requirements
    -0.60
     mandatory
    -0.59
    Minor
    -0.59
     consecut
    -0.55
     filler
    -0.55
     consecutive
    -0.55
     decomp
    -0.55
     bases
    -0.55
    POSITIVE LOGITS
    conom
    0.75
    ãĥĬ
    0.68
     JPM
    0.66
    NYSE
    0.64
     futures
    0.64
     Mits
    0.63
    fortune
    0.63
     Aph
    0.60
    aj
    0.60
     yen
    0.60
    Activations Density 0.011%

    No Known Activations