© 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 12288
    9. 8-RES_FS12288-JB
    11. 10098
    Prev
    Next
    INDEX
    Explanations

    verbs related to physical action and movement, particularly the action of "getting."

    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
    314
    +0.20
    0.9%
    566
    +0.10
    0.5%
    183
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    183
    +0.20
    0.06
    282
    +0.10
    0.06
    566
    +0.10
    0.05
    Negative Logits
     Lies
    -0.66
     cardinal
    -0.64
     Archdemon
    -0.62
     experiment
    -0.61
     Britann
    -0.60
     Inher
    -0.60
     Wonders
    -0.60
    iege
    -0.59
     Tanz
    -0.57
     Fail
    -0.57
    POSITIVE LOGITS
     rid
    1.57
     acquainted
    1.08
    TING
    0.91
     distracted
    0.91
     away
    0.90
    bent
    0.89
    cloneembedreportprint
    0.88
    tin
    0.87
    away
    0.86
     Started
    0.84
    Activations Density 0.132%

    No Known Activations