© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 7323
    Prev
    Next
    INDEX
    Explanations

    phrases related to making responsible decisions in complex situations

    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
    447
    +0.21
    0.9%
    358
    +0.11
    0.5%
    626
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    737
    +0.21
    0.04
    247
    +0.11
    0.04
    666
    +0.11
    0.04
    Negative Logits
    ioxide
    -0.77
    iciency
    -0.69
    oat
    -0.66
    arching
    -0.65
    ixture
    -0.64
    inite
    -0.64
    ufficient
    -0.63
    ongo
    -0.63
    amation
    -0.62
    ater
    -0.61
    POSITIVE LOGITS
     athlet
    0.79
     wisely
    0.77
     hoping
    0.77
     whereby
    0.77
     offensively
    0.75
     defensively
    0.75
     gamble
    0.75
     unilaterally
    0.74
     boldly
    0.72
     regarding
    0.72
    Activations Density 0.225%

    No Known Activations