Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 23919
    Prev
    Next
    INDEX
    Explanations

    specific nouns related to bureaucratic processes and structures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    611
    +0.13
    0.6%
    438
    +0.12
    0.5%
    313
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.07
    575
    +0.12
    0.05
    438
    +0.10
    0.04
    Negative Logits
    ervative
    -0.79
    creen
    -0.78
    risome
    -0.70
    cale
    -0.69
    enza
    -0.69
    elcome
    -0.68
    hetto
    -0.67
    ilk
    -0.65
    ritz
    -0.65
    atom
    -0.65
    POSITIVE LOGITS
    lets
    0.93
    ologies
    0.86
    naires
    0.84
    cases
    0.80
    backs
    0.79
    worms
    0.79
    ages
    0.76
     proportions
    0.76
    lies
    0.75
    holders
    0.74
    Activations Density 0.316%

    No Known Activations