© 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. 853
    Prev
    Next
    INDEX
    Explanations

    phrases related to organization, management, development, maintenance, operation, trade, capture, and record in a technical context

    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
    504
    +0.11
    0.5%
    79
    +0.11
    0.5%
    560
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    70
    +0.11
    0.05
    79
    +0.11
    0.05
    505
    +0.09
    0.04
    Negative Logits
    atus
    -0.67
    antha
    -0.67
    liest
    -0.67
    culus
    -0.66
    usra
    -0.64
    voy
    -0.63
    imura
    -0.63
    cus
    -0.62
    icus
    -0.62
    cas
    -0.62
    POSITIVE LOGITS
     alike
    1.24
     respectively
    1.00
     accordingly
    0.78
    enance
    0.66
     portions
    0.63
     aspects
    0.61
     depending
    0.60
     amongst
    0.60
    depending
    0.59
     oneself
    0.59
    Activations Density 0.355%

    No Known Activations