© 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 EvalsExportsinterp-engineNEWAPI 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. 16746
    Prev
    Next
    INDEX
    Explanations

    positive and negative reactions or emotions to political figures and events

    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
    426
    +0.11
    0.5%
    720
    +0.10
    0.5%
    378
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.11
    0.08
    547
    +0.10
    0.04
    526
    +0.09
    -0.01
    Negative Logits
    agate
    -0.76
     planner
    -0.71
     exercises
    -0.70
     Parallel
    -0.67
    artifacts
    -0.65
    empl
    -0.63
     supplements
    -0.62
    wine
    -0.62
    otta
    -0.60
    walking
    -0.59
    POSITIVE LOGITS
     audiences
    1.27
     onlook
    1.14
     detractors
    1.14
     critics
    1.08
     applause
    1.03
     listeners
    1.02
     viewers
    1.02
     audience
    0.98
     crowds
    0.97
     skeptics
    0.97
    Activations Density 0.966%

    No Known Activations