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. 1134
    Prev
    Next
    INDEX
    Explanations

    phrases related to holding onto beliefs, power, responsibilities, or physical objects

    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
    346
    +0.11
    0.5%
    582
    +0.11
    0.5%
    183
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    346
    +0.11
    0.05
    582
    +0.11
    0.04
    531
    +0.10
    0.04
    Negative Logits
    ————
    -0.70
    DIT
    -0.66
    nown
    -0.66
    ãĤ¡
    -0.65
    ixel
    -0.64
    gnu
    -0.64
    ibel
    -0.64
    ghan
    -0.64
    ettel
    -0.64
    endix
    -0.63
    POSITIVE LOGITS
    hold
    1.09
     sway
    1.04
    erness
    0.98
     accountable
    0.97
    holding
    0.97
    holders
    0.95
     hostage
    0.92
     captive
    0.88
    holder
    0.88
     onto
    0.87
    Activations Density 0.491%

    No Known Activations