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. 2368
    Prev
    Next
    INDEX
    Explanations

    adverbs and adjectives that describe frequency or probability

    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
    208
    +0.10
    0.4%
    447
    +0.09
    0.4%
    630
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    208
    +0.10
    0.06
    724
    +0.09
    0.07
    151
    +0.08
    0.07
    Negative Logits
    anth
    -0.70
     Kob
    -0.68
    eday
    -0.68
    yet
    -0.64
     Integrity
    -0.63
     Orchestra
    -0.62
    eem
    -0.62
    hetically
    -0.60
     Push
    -0.59
     Kah
    -0.58
    POSITIVE LOGITS
    entimes
    1.04
     speaking
    0.88
     overlooked
    0.86
     abbrevi
    0.83
     regarded
    0.81
     frowned
    0.79
     reserved
    0.79
     consist
    0.78
     consists
    0.76
     associated
    0.76
    Activations Density 0.635%

    No Known Activations