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 49152
    5. 8-RES_FS49152-JB
    6. 46174
    Prev
    Next
    INDEX
    Explanations

    instances of conversational or social interaction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    756
    +0.10
    0.4%
    250
    +0.10
    0.4%
    493
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    493
    +0.10
    0.03
    70
    +0.10
    0.03
    196
    +0.10
    0.03
    Negative Logits
    fits
    -0.65
    mega
    -0.64
    ulo
    -0.62
    allows
    -0.62
    }"
    -0.61
    iru
    -0.61
    Definition
    -0.59
     Publication
    -0.58
     Flavoring
    -0.58
    }:
    -0.58
    POSITIVE LOGITS
     about
    0.84
     aloud
    0.82
     backstage
    0.76
     briefly
    0.74
     amongst
    0.73
     acquaintances
    0.71
     banter
    0.71
     frankly
    0.71
     Skype
    0.70
     discussing
    0.69
    Activations Density 0.047%

    No Known Activations