© 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 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. 10365
    Prev
    Next
    INDEX
    Explanations

    The main thing this neuron does is to find parts of text that indicate the continuation of reading in an article

    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
    373
    +0.14
    0.7%
    257
    +0.10
    0.5%
    293
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    257
    +0.14
    0.02
    530
    +0.10
    0.02
    28
    +0.08
    0.02
    Negative Logits
    »Ĵ
    -0.81
     Gleaming
    -0.69
    ignt
    -0.67
    footed
    -0.62
    folk
    -0.61
    è¦ļéĨĴ
    -0.61
    ĪĴ
    -0.59
    icion
    -0.59
    ouls
    -0.59
    vir
    -0.58
    POSITIVE LOGITS
     Below
    0.70
     isEnabled
    0.65
    ...]
    0.63
     aloud
    0.61
     toggle
    0.60
     ARTICLE
    0.59
     POLITICO
    0.59
     BELOW
    0.59
     âĨĴ
    0.57
     ETH
    0.56
    Activations Density 0.012%

    No Known Activations