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. 26072
    Prev
    Next
    INDEX
    Explanations

    quantities and values related to money and transactions

    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
    409
    +0.12
    0.5%
    589
    +0.11
    0.5%
    577
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    409
    +0.12
    0.02
    767
    +0.11
    0.02
    407
    +0.10
    0.01
    Negative Logits
     Rosenberg
    -0.85
    Adapter
    -0.84
     attorneys
    -0.75
     Piper
    -0.74
     Center
    -0.72
     Proposition
    -0.72
     Cosponsors
    -0.71
     Arkansas
    -0.71
    inois
    -0.70
    paio
    -0.70
    POSITIVE LOGITS
     £
    1.44
    £
    1.39
     (£
    1.21
    UK
    1.08
     realise
    1.07
    iverpool
    1.04
     labou
    1.04
    ´
    1.02
     Britain
    1.01
    Britain
    1.00
    Activations Density 0.078%

    No Known Activations