© 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. 10401
    Prev
    Next
    INDEX
    Explanations

    legal terms related to legal accusations and procedures

    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
    447
    +0.14
    0.7%
    679
    +0.12
    0.6%
    288
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    100
    +0.14
    0.02
    540
    +0.12
    0.02
    203
    +0.11
    0.01
    Negative Logits
    birth
    -0.87
     lear
    -0.72
    atomic
    -0.70
    istically
    -0.70
    conservancy
    -0.69
    istic
    -0.68
    atan
    -0.65
     Kers
    -0.65
    away
    -0.65
     quarters
    -0.63
    POSITIVE LOGITS
     indict
    1.22
     indictment
    1.18
     indicted
    1.13
    eering
    1.06
     accuses
    0.90
     alleges
    0.85
     wrongdoing
    0.85
     accused
    0.84
     accuse
    0.84
     warrant
    0.83
    Activations Density 0.032%

    No Known Activations