© 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 49152
    9. 8-RES_FS49152-JB
    11. 41492
    Prev
    Next
    INDEX
    Explanations

    terms related to searching or investigations

    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
    129
    +0.10
    0.5%
    72
    +0.09
    0.4%
    531
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    129
    +0.10
    0.02
    97
    +0.09
    0.01
    263
    +0.09
    0.02
    Negative Logits
    ivari
    -0.76
    Production
    -0.75
     Leilan
    -0.67
    nect
    -0.65
    minecraft
    -0.65
    noxious
    -0.65
    puff
    -0.63
    terness
    -0.63
     cultured
    -0.63
    Earn
    -0.62
    POSITIVE LOGITS
     warrant
    1.05
     Warrant
    0.96
     warrants
    0.94
    lights
    0.89
    able
    0.83
     engines
    0.83
     engine
    0.79
    light
    0.74
    alysed
    0.72
    ability
    0.71
    Activations Density 0.014%

    No Known Activations