© 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. 12127
    Prev
    Next
    INDEX
    Explanations

    dialogue cues and descriptions related to characters in a story

    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
    480
    +0.49
    2.6%
    447
    +0.16
    0.9%
    266
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.49
    0.02
    266
    +0.16
    0.07
    590
    +0.14
    0.04
    Negative Logits
     promoters
    -0.63
    conservancy
    -0.60
     detriment
    -0.59
    avorite
    -0.58
    Sources
    -0.58
    uilt
    -0.57
    halla
    -0.57
    senal
    -0.57
     redes
    -0.56
     IOC
    -0.56
    POSITIVE LOGITS
     smir
    0.93
     startled
    0.91
     softly
    0.91
     nodded
    0.91
     nervously
    0.90
     sigh
    0.88
     grinned
    0.87
     grin
    0.87
     murm
    0.87
     trembling
    0.87
    Activations Density 0.706%

    No Known Activations