© 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 12288
    9. 8-RES_FS12288-JB
    11. 1532
    Prev
    Next
    INDEX
    Explanations

    the word "yet" used in specific contexts

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    757
    +0.12
    0.5%
    312
    +0.10
    0.4%
    709
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    135
    +0.12
    0.03
    8
    +0.10
    0.03
    757
    +0.10
    0.02
    Negative Logits
    hal
    -0.75
    ga
    -0.75
    tained
    -0.74
    tein
    -0.74
    ding
    -0.70
    ãĤ¦ãĤ¹
    -0.70
    gang
    -0.70
    ãĤ¡
    -0.69
    ãĥİ
    -0.68
    packs
    -0.68
    POSITIVE LOGITS
     somehow
    1.14
     strangely
    0.91
     somew
    0.84
     persisted
    0.81
     despite
    0.81
     nonetheless
    0.80
     persists
    0.79
     mirac
    0.77
    entimes
    0.74
     interestingly
    0.74
    Activations Density 0.046%

    No Known Activations