© 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 24576
    9. 8-RES_FS24576-JB
    11. 18330
    Prev
    Next
    INDEX
    Explanations

    social media posts related to art and design

    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
    481
    +0.22
    1.0%
    745
    +0.14
    0.6%
    671
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.22
    0.04
    442
    +0.14
    0.03
    257
    +0.11
    0.02
    Negative Logits
     Adin
    -0.75
     antidepressant
    -0.73
     prolong
    -0.69
     Kling
    -0.67
     retard
    -0.65
     enriched
    -0.65
     arithmetic
    -0.64
     infants
    -0.64
     elephants
    -0.63
     recount
    -0.63
    POSITIVE LOGITS
    www
    1.30
    http
    1.24
    https
    1.22
    ttp
    1.19
    web
    1.17
    blogspot
    1.14
    Website
    1.14
    cdn
    1.13
    twitter
    1.10
    wiki
    1.06
    Activations Density 0.088%

    No Known Activations