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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1594
    Prev
    Next
    INDEX
    Explanations

    problems or challenges and their solutions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    35
    +0.13
    0.6%
    211
    +0.10
    0.5%
    198
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    51
    +0.13
    0.07
    35
    +0.10
    0.06
    211
    +0.10
    0.03
    Negative Logits
    rib
    -0.83
    urses
    -0.78
    alty
    -0.76
    rica
    -0.76
    robe
    -0.74
    vez
    -0.74
    ensions
    -0.73
    orks
    -0.73
    eston
    -0.71
    ron
    -0.71
    POSITIVE LOGITS
     plag
    1.41
     solved
    1.03
     solving
    0.98
    hooting
    0.96
     facing
    0.89
     confronting
    0.88
    Problem
    0.86
     stemmed
    0.86
     stemming
    0.86
     arise
    0.85
    Activations Density 0.750%

    No Known Activations