© 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 12288
    9. 8-RES_FS12288-JB
    11. 10419
    Prev
    Next
    INDEX
    Explanations

    phrases related to police violence and societal issues, as well as scientific research and manipulation of public opinion

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    87
    +0.24
    1.2%
    756
    +0.12
    0.6%
    741
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    756
    +0.24
    -0.03
    87
    +0.12
    -0.04
    526
    +0.09
    -0.02
    Negative Logits
     Finder
    -0.80
    Berry
    -0.71
    estamp
    -0.68
    anse
    -0.65
    reci
    -0.65
    appiness
    -0.64
    availability
    -0.64
     Observ
    -0.63
    Enlarge
    -0.63
    estones
    -0.63
    POSITIVE LOGITS
     scapego
    1.21
     shameless
    1.04
     arrog
    0.98
     incompetence
    0.98
     insidious
    0.97
     phony
    0.95
     blackmail
    0.95
     ignor
    0.95
     extortion
    0.94
     fools
    0.94
    Activations Density 2.054%

    No Known Activations