© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 37311
    Prev
    Next
    INDEX
    Explanations

    patterns related to data structures and array-like syntaxes in programming

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    701
    +0.11
    0.5%
    348
    +0.11
    0.5%
    112
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    701
    +0.11
    0.01
    160
    +0.11
    0.01
    364
    +0.11
    0.01
    Negative Logits
    otted
    -0.78
    hered
    -0.75
    wagen
    -0.74
     seams
    -0.71
    iquette
    -0.66
     unden
    -0.66
     nodd
    -0.66
     proport
    -0.65
     lapt
    -0.63
    jug
    -0.63
    POSITIVE LOGITS
    ?]
    1.31
    ],
    1.27
    ](
    1.25
    email
    1.19
    !]
    1.18
    ])
    1.18
    ].
    1.17
    ...]
    1.16
    ][
    1.13
    actionDate
    1.11
    Activations Density 0.009%

    No Known Activations