© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 9232
    Prev
    Next
    INDEX
    Explanations

    information, especially details or insider information

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    447
    +0.26
    1.2%
    119
    +0.12
    0.5%
    495
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    119
    +0.26
    0.05
    495
    +0.12
    0.05
    591
    +0.11
    0.04
    Negative Logits
    ortunately
    -0.68
    igham
    -0.66
    emale
    -0.64
     Roses
    -0.64
    xtap
    -0.63
    ategory
    -0.61
     Carbuncle
    -0.60
    Zone
    -0.60
     Balanced
    -0.60
    ovie
    -0.60
    POSITIVE LOGITS
     regarding
    1.15
     relating
    1.04
     pertaining
    1.03
     about
    1.03
     concerning
    1.02
     glean
    1.00
     detailing
    0.93
     ABOUT
    0.85
     briefings
    0.82
     clues
    0.80
    Activations Density 0.174%

    No Known Activations