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 3072
    5. 8-RES_FS3072-JB
    6. 393
    Prev
    Next
    INDEX
    Explanations

    terms related to lists or listings

    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
    558
    +0.12
    0.6%
    39
    +0.11
    0.5%
    222
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    39
    +0.12
    0.06
    558
    +0.11
    0.06
    222
    +0.10
    0.05
    Negative Logits
     Aber
    -0.73
     Huck
    -0.66
     Gore
    -0.58
     Pagan
    -0.58
     Galile
    -0.57
     Ao
    -0.56
     Advocate
    -0.56
     Ath
    -0.56
     Hai
    -0.56
    irgin
    -0.56
    POSITIVE LOGITS
    erv
    1.12
    ening
    0.95
    lists
    0.88
    icles
    0.85
    erve
    0.84
    icter
    0.84
    icle
    0.84
     listing
    0.83
    ener
    0.81
    eners
    0.81
    Activations Density 0.762%

    No Known Activations