© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 6555
    Prev
    Next
    INDEX
    Explanations

    statements asserting the safety or assurance of something

    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
    682
    +0.12
    0.5%
    416
    +0.10
    0.5%
    481
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    189
    +0.12
    0.03
    69
    +0.10
    0.03
    528
    +0.10
    0.03
    Negative Logits
     CrossRef
    -0.77
     unfocusedRange
    -0.67
     Springer
    -0.67
    Suddenly
    -0.65
    Joined
    -0.64
    igslist
    -0.60
     goose
    -0.59
     hilar
    -0.56
    pmwiki
    -0.55
     suddenly
    -0.55
    POSITIVE LOGITS
     innocence
    1.04
     abide
    0.89
    absolutely
    0.80
     comply
    0.79
     neutrality
    0.79
     abiding
    0.78
     confidentiality
    0.77
     nothing
    0.76
     complied
    0.76
    never
    0.75
    Activations Density 0.353%

    No Known Activations