© 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
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 19950
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to returning to a previous point or past occurrences

    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
    377
    +0.12
    0.5%
    659
    +0.12
    0.5%
    50
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    50
    +0.12
    0.03
    143
    +0.12
    0.02
    488
    +0.11
    0.02
    Negative Logits
     constitu
    -0.66
    izo
    -0.65
     tyr
    -0.63
     Osc
    -0.62
     agre
    -0.61
    inational
    -0.60
    ciation
    -0.60
    aukee
    -0.59
    riages
    -0.58
    æĪ¦
    -0.58
    POSITIVE LOGITS
    lash
    1.15
    GROUND
    1.11
    stab
    1.05
    ward
    1.01
    tracking
    0.99
    door
    0.99
    wards
    0.98
    story
    0.95
    packs
    0.95
    pack
    0.94
    Activations Density 0.021%

    No Known Activations