© 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 49152
    9. 8-RES_FS49152-JB
    11. 37330
    Prev
    Next
    INDEX
    Explanations

    numerical values or statistics related to various subjects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    314
    +0.12
    0.5%
    373
    +0.10
    0.4%
    76
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    632
    +0.12
    0.02
    742
    +0.10
    0.01
    357
    +0.09
    0.01
    Negative Logits
     toile
    -0.72
    humans
    -0.64
     forth
    -0.63
    umerable
    -0.62
     scripting
    -0.61
    forth
    -0.60
     questioning
    -0.58
     lodging
    -0.56
     authors
    -0.56
    fters
    -0.56
    POSITIVE LOGITS
    ufact
    1.02
    gage
    0.92
    icio
    0.85
    ONEY
    0.77
    ascar
    0.76
    udd
    0.76
    ICAN
    0.75
    ains
    0.72
    arijuana
    0.70
    achine
    0.70
    Activations Density 0.016%

    No Known Activations