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. 2759
    Prev
    Next
    INDEX
    Explanations

    verbs related to providing or offering

    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
    59
    +0.13
    0.6%
    642
    +0.12
    0.5%
    421
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    59
    +0.13
    0.06
    642
    +0.12
    0.05
    761
    +0.10
    0.05
    Negative Logits
    doms
    -0.71
     Nanto
    -0.68
    cus
    -0.68
    phase
    -0.68
    brow
    -0.66
    mat
    -0.66
    beat
    -0.65
    walker
    -0.64
    dar
    -0.64
    war
    -0.64
    POSITIVE LOGITS
     ample
    0.96
     sust
    0.92
    utical
    0.92
     guidance
    0.89
     logistical
    0.89
     insight
    0.88
     assistance
    0.88
     refunds
    0.88
     adequate
    0.86
     invaluable
    0.82
    Activations Density 0.506%

    No Known Activations