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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 3072
    9. 8-RES_FS3072-JB
    11. 744
    Prev
    Next
    INDEX
    Explanations

    information related to personal experiences and stories

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    217
    +0.13
    0.6%
    640
    +0.12
    0.6%
    373
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.13
    0.13
    6
    +0.12
    0.11
    160
    +0.10
    0.08
    Negative Logits
     stret
    -0.73
     downstream
    -0.70
     synerg
    -0.70
     funeral
    -0.65
     restless
    -0.65
     migr
    -0.64
     weaving
    -0.64
     vaccinated
    -0.64
     hust
    -0.64
     netted
    -0.63
    POSITIVE LOGITS
    Answer
    1.86
     Answer
    1.16
    Yeah
    1.11
    Well
    1.06
    Absolutely
    1.06
    Yes
    1.05
    ANS
    1.01
    answer
    0.99
    RH
    0.96
    Solution
    0.95
    Activations Density 1.168%

    No Known Activations