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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 37099
    Prev
    Next
    INDEX
    Explanations

    words related to legal consequences or conditions

    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
    461
    +0.13
    0.6%
    642
    +0.10
    0.5%
    330
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    642
    +0.13
    0.01
    461
    +0.10
    0.01
    224
    +0.10
    0.01
    Negative Logits
    iffe
    -0.76
    isoft
    -0.67
    imet
    -0.61
    enic
    -0.61
     Carbuncle
    -0.61
    ortun
    -0.60
     Neighbor
    -0.60
    ãĤ¼
    -0.59
    anqu
    -0.59
    ainted
    -0.58
    POSITIVE LOGITS
     unequivocally
    1.07
     bluntly
    0.99
     emphatically
    0.92
     plainly
    0.91
     goodbye
    0.79
     categor
    0.75
     boldly
    0.74
     quo
    0.74
     "...
    0.74
     confidently
    0.73
    Activations Density 0.011%

    No Known Activations