© 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 1536
    9. 8-RES_FS1536-JB
    11. 13
    Prev
    Next
    INDEX
    Explanations

    information relating to the Real IRA's actions and statements

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    373
    +0.21
    1.3%
    314
    +0.08
    0.5%
    516
    +0.08
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.21
    -0.24
    298
    +0.08
    0.10
    620
    +0.08
    -0.01
    Negative Logits
     [+
    -0.39
     Originally
    -0.37
    iple
    -0.36
    ÃŁ
    -0.35
     *)
    -0.35
    Å¡
    -0.35
     Referred
    -0.34
    \)
    -0.34
    Äį
    -0.34
    aez
    -0.34
    POSITIVE LOGITS
    '.
    0.55
     theirs
    0.54
    %.
    0.53
    $.
    0.53
     attRot
    0.48
    ".
    0.47
    *.
    0.47
    etheless
    0.46
    ãĤ´ãĥ³
    0.46
    .''
    0.44
    Activations Density 13.959%

    No Known Activations