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. 40096
    Prev
    Next
    INDEX
    Explanations

    phrases indicating levels of intensity or comparisons

    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
    718
    +0.10
    0.5%
    741
    +0.09
    0.4%
    466
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.04
    222
    +0.09
    0.03
    362
    +0.09
    0.04
    Negative Logits
    assies
    -0.84
    paces
    -0.80
    ravings
    -0.80
    pages
    -0.79
    arenthood
    -0.78
    ographics
    -0.77
    otions
    -0.74
    rollers
    -0.73
    events
    -0.73
    runs
    -0.72
    POSITIVE LOGITS
     euphem
    1.13
     reminder
    1.12
     hallmark
    1.09
     precursor
    1.05
     staple
    1.04
     indication
    1.02
     testament
    1.01
     tactic
    0.97
     impossibility
    0.97
     indicator
    0.95
    Activations Density 0.226%

    No Known Activations