© 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. 10727
    Prev
    Next
    INDEX
    Explanations

    phrases related to expressing high levels of confidence in a statement

    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
    4
    +0.10
    0.5%
    196
    +0.10
    0.4%
    462
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    676
    +0.10
    0.03
    208
    +0.10
    0.03
    196
    +0.10
    0.03
    Negative Logits
    worldly
    -0.89
    zie
    -0.81
    isson
    -0.73
    isure
    -0.71
    raltar
    -0.70
    oga
    -0.70
    andr
    -0.68
    EStream
    -0.68
    pload
    -0.68
    uckland
    -0.68
    POSITIVE LOGITS
     Invisible
    0.76
     accuracy
    0.74
     certainty
    0.70
     accurate
    0.70
     authentic
    0.70
     agree
    0.67
     pure
    0.66
    rated
    0.64
     authenticated
    0.64
     ethanol
    0.64
    Activations Density 0.060%

    No Known Activations