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

    Neuronpedia

    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. 24060
    Prev
    Next
    INDEX
    Explanations

    references to editorial actions or annotations

    oai_token-act-pair · gpt-4o-miniTriggered 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
    373
    +0.14
    0.6%
    455
    +0.11
    0.5%
    555
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.14
    0.04
    555
    +0.11
    0.04
    358
    +0.10
    0.04
    Negative Logits
    milo
    -0.87
    ciating
    -0.71
     pige
    -0.66
    dinand
    -0.65
     advoc
    -0.65
     pigeon
    -0.64
     mosqu
    -0.63
    emouth
    -0.61
     Engineers
    -0.60
    mercial
    -0.58
    POSITIVE LOGITS
     ]
    1.24
     ],
    0.94
     ])
    0.92
     ].
    0.86
     .)
    0.80
     edit
    0.78
     ][
    0.77
     )
    0.74
    Spoiler
    0.72
    ī
    0.71
    Activations Density 0.013%

    No Known Activations