© 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. 41713
    Prev
    Next
    INDEX
    Explanations

    references to specific financial instruments or trading activities

    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
    167
    +0.14
    0.6%
    146
    +0.12
    0.6%
    743
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.14
    0.01
    168
    +0.12
    0.01
    326
    +0.10
    0.01
    Negative Logits
    perse
    -0.65
    housing
    -0.64
     Pry
    -0.61
    etheless
    -0.61
    ÄŁ
    -0.60
    ength
    -0.59
    ogue
    -0.59
     ward
    -0.59
    cki
    -0.58
     kHz
    -0.58
    POSITIVE LOGITS
    allery
    0.97
    ITAL
    0.97
    arette
    0.95
    arettes
    0.95
    gers
    0.94
    GER
    0.93
    RANT
    0.93
    INAL
    0.91
    REE
    0.91
    rams
    0.89
    Activations Density 0.009%

    No Known Activations