© 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 3072
    9. 8-RES_FS3072-JB
    11. 707
    Prev
    Next
    INDEX
    Explanations

    words related to actions and consequences, especially in legal or authority 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_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    357
    +0.09
    0.4%
    312
    +0.09
    0.4%
    503
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    312
    +0.09
    0.06
    655
    +0.09
    0.07
    384
    +0.09
    0.05
    Negative Logits
    igger
    -0.82
    owler
    -0.75
    FactoryReloaded
    -0.72
    actionDate
    -0.71
    resa
    -0.70
    ELD
    -0.70
    amba
    -0.70
    ibble
    -0.69
    agger
    -0.69
    Slot
    -0.67
    POSITIVE LOGITS
     vice
    0.84
     proceeded
    0.82
     thereafter
    0.81
     expects
    0.79
     incorporates
    0.78
     guessed
    0.76
     invests
    0.76
     intends
    0.75
     succumb
    0.75
     froze
    0.75
    Activations Density 1.185%

    No Known Activations