© 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. 2267
    Prev
    Next
    INDEX
    Explanations

    legal terms and analysis related to a specific legal case

    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
    743
    +0.11
    0.5%
    640
    +0.10
    0.4%
    160
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.11
    0.07
    160
    +0.10
    0.05
    743
    +0.09
    0.03
    Negative Logits
     nightly
    -0.72
     gigs
    -0.70
     hug
    -0.70
     hardcore
    -0.69
     passionate
    -0.68
     branching
    -0.68
     angry
    -0.67
     deep
    -0.65
     rooft
    -0.65
     chem
    -0.65
    POSITIVE LOGITS
    §
    1.33
    shall
    1.32
    Section
    1.27
     ``(
    1.15
    Effective
    1.15
     Notwithstanding
    1.15
    Marginal
    1.14
    Applic
    1.11
    Definition
    1.11
    Exception
    1.09
    Activations Density 0.200%

    No Known Activations