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 24576
    5. 8-RES_FS24576-JB
    6. 23100
    Prev
    Next
    INDEX
    Explanations

    phrases indicating movement away from something

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    113
    +0.13
    0.6%
    314
    +0.13
    0.6%
    513
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    113
    +0.13
    0.05
    347
    +0.13
    0.04
    693
    +0.10
    0.04
    Negative Logits
     mosaic
    -0.66
    Else
    -0.64
    Sequ
    -0.63
     alloc
    -0.62
     warrants
    -0.61
    EMS
    -0.57
    IRE
    -0.57
     superflu
    -0.57
     basis
    -0.56
     SVG
    -0.56
    POSITIVE LOGITS
    enced
    0.88
    fed
    0.84
    raged
    0.80
    oing
    0.78
    doing
    0.77
    loaded
    0.75
    joy
    0.72
     ready
    0.72
    locked
    0.70
    joice
    0.70
    Activations Density 0.209%

    No Known Activations