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 768
    5. 8-RES_FS768-JB
    6. 271
    Prev
    Next
    INDEX
    Explanations

    economic and political discussion related to blame assignment and social issues

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    480
    +0.74
    5.5%
    266
    +0.16
    1.2%
    481
    +0.14
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.74
    0.04
    266
    +0.16
    0.41
    351
    +0.14
    0.12
    Negative Logits
     Canaver
    -0.47
    iannopoulos
    -0.42
     Tsarnaev
    -0.42
    ãĤ¨ãĥ«
    -0.42
    ihadi
    -0.40
    annabin
    -0.38
    ovember
    -0.38
    Liberal
    -0.38
    realDonaldTrump
    -0.37
     Fresno
    -0.37
    POSITIVE LOGITS
     alot
    0.44
     ones
    0.44
     tho
    0.40
    @#&
    0.40
    )</
    0.40
     afterwards
    0.38
     usually
    0.38
    )--
    0.38
     google
    0.38
    !),
    0.38
    Activations Density 19.956%

    No Known Activations