Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 110
    Prev
    Next
    INDEX
    Explanations

    discussions related to political and economic systems

    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
    308
    +0.12
    0.6%
    290
    +0.10
    0.5%
    314
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    308
    +0.12
    0.10
    720
    +0.10
    0.10
    546
    +0.10
    0.09
    Negative Logits
    orts
    -0.75
    icable
    -0.65
    ousel
    -0.63
     commun
    -0.62
    eks
    -0.62
    iple
    -0.60
    ruck
    -0.60
    oyd
    -0.60
     level
    -0.59
     trainer
    -0.59
    POSITIVE LOGITS
    _-
    1.23
    perhaps
    1.17
    especially
    1.17
    ————
    1.15
    particularly
    1.12
    ————————
    1.09
    including
    1.07
    something
    1.05
    albeit
    1.03
    that
    1.01
    Activations Density 1.900%

    No Known Activations