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 1536
    5. 8-RES_FS1536-JB
    6. 604
    Prev
    Next
    INDEX
    Explanations

    phrases related to societal issues and differences

    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
    408
    +0.11
    0.5%
    395
    +0.11
    0.5%
    7
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    126
    +0.11
    0.10
    395
    +0.11
    0.10
    408
    +0.11
    0.13
    Negative Logits
    assisted
    -0.74
    buster
    -0.68
    his
    -0.67
    ãĥ¼ãĤ¯
    -0.66
    ares
    -0.65
    ilk
    -0.64
    aves
    -0.64
    late
    -0.63
    ires
    -0.63
    union
    -0.63
    POSITIVE LOGITS
     overlap
    1.10
     shortage
    1.00
     difference
    1.00
     possibility
    0.98
     waiting
    0.97
     downside
    0.95
     inherent
    0.95
     reason
    0.95
     lurking
    0.94
     discrepancy
    0.93
    Activations Density 2.159%

    No Known Activations