© 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. 1784
    Prev
    Next
    INDEX
    Explanations

    information related to regulations and guidelines, particularly in the context of travel services, education, and career counseling for veterans

    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
    640
    +0.21
    1.0%
    481
    +0.18
    0.8%
    447
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.21
    0.13
    6
    +0.18
    0.08
    160
    +0.12
    0.08
    Negative Logits
     damned
    -0.91
     Frankenstein
    -0.86
     banging
    -0.83
     grinning
    -0.82
     damn
    -0.81
     overpower
    -0.80
     overpowered
    -0.80
     dudes
    -0.80
     fuck
    -0.78
     smashing
    -0.78
    POSITIVE LOGITS
    Information
    1.28
    Services
    1.23
    Learn
    1.20
    Contact
    1.18
    Types
    1.16
    Registration
    1.12
    FAQ
    1.10
     Availability
    1.10
     Regulations
    1.07
    Important
    1.06
    Activations Density 0.448%

    No Known Activations