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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 4946
    Prev
    Next
    INDEX
    Explanations

    financial terms related to stock markets, indices, companies, and economic indicators

    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
    167
    +0.12
    0.5%
    326
    +0.11
    0.5%
    0
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.01
    0
    +0.11
    0.01
    99
    +0.10
    0.01
    Negative Logits
    REDACTED
    -0.77
    é»Ĵ
    -0.74
    ologies
    -0.67
    opathic
    -0.66
     Reach
    -0.60
    ACTED
    -0.59
     extradition
    -0.59
    opath
    -0.59
     Wanted
    -0.59
     Mongolia
    -0.59
    POSITIVE LOGITS
    lings
    1.02
    minster
    1.01
    cake
    0.98
    ager
    0.93
    ishes
    0.91
    rey
    0.89
    ishers
    0.88
    ishing
    0.87
    ry
    0.86
    ling
    0.86
    Activations Density 0.020%

    No Known Activations