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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 23838
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate pathways or opportunities for future developments or actions

    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
    397
    +0.12
    0.5%
    202
    +0.12
    0.5%
    236
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    397
    +0.12
    0.01
    236
    +0.12
    0.01
    737
    +0.11
    0.01
    Negative Logits
    iosyncr
    -0.67
    rities
    -0.66
    inately
    -0.66
    kov
    -0.64
     brill
    -0.62
    itu
    -0.61
    hemat
    -0.61
     pecul
    -0.60
     rag
    -0.60
    sports
    -0.60
    POSITIVE LOGITS
    ward
    0.87
     toward
    0.86
     ahead
    0.81
     towards
    0.81
    forward
    0.79
    WARD
    0.78
    fare
    0.75
     forward
    0.74
    ways
    0.74
     paved
    0.71
    Activations Density 0.012%

    No Known Activations