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 49152
    5. 8-RES_FS49152-JB
    6. 26410
    Prev
    Next
    INDEX
    Explanations

    mentions of monetary support or contributions

    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
    302
    +0.11
    0.5%
    132
    +0.10
    0.5%
    699
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    302
    +0.11
    0.01
    13
    +0.10
    0.01
    156
    +0.10
    0.01
    Negative Logits
    Fall
    -0.69
     Ferguson
    -0.69
    pmwiki
    -0.69
    Thumbnail
    -0.66
    iom
    -0.64
    Leaks
    -0.64
    pex
    -0.63
     Azerb
    -0.61
     Nile
    -0.58
     Moff
    -0.58
    POSITIVE LOGITS
     inclusive
    0.66
     reserved
    0.64
     Tradable
    0.63
     advanced
    0.61
     cups
    0.59
     applicable
    0.59
     importantly
    0.58
     pounds
    0.58
     sponsoring
    0.58
     certs
    0.58
    Activations Density 0.006%

    No Known Activations