© 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 1536
    9. 8-RES_FS1536-JB
    11. 59
    Prev
    Next
    INDEX
    Explanations

    instances of the word "see" followed by other words or phrases

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    240
    +0.11
    0.5%
    395
    +0.11
    0.5%
    218
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    395
    +0.11
    0.08
    240
    +0.11
    0.07
    403
    +0.10
    0.07
    Negative Logits
    inqu
    -0.71
    onto
    -0.69
    trap
    -0.68
    ÃŃa
    -0.67
     whistle
    -0.66
    eware
    -0.65
    externalActionCode
    -0.64
    nesia
    -0.64
     agric
    -0.64
    prime
    -0.63
    POSITIVE LOGITS
    dust
    1.13
     firsthand
    0.99
     visions
    0.91
     parallels
    0.88
    awed
    0.83
    lights
    0.82
     similarities
    0.80
     eye
    0.79
     glimps
    0.77
     fit
    0.76
    Activations Density 1.205%

    No Known Activations