© 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 24576
    9. 8-RES_FS24576-JB
    11. 7407
    Prev
    Next
    INDEX
    Explanations

    phrases related to the significance or importance of various matters

    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
    739
    +0.10
    0.5%
    49
    +0.10
    0.5%
    431
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.10
    0.02
    431
    +0.10
    0.02
    125
    +0.10
    0.02
    Negative Logits
    ãĥ³ãĤ¸
    -0.81
    ahime
    -0.79
    ©¶æ
    -0.79
    rys
    -0.76
    ARP
    -0.74
    guyen
    -0.74
    uthor
    -0.68
    oute
    -0.68
    apons
    -0.68
    ãĥĩãĤ£
    -0.64
    POSITIVE LOGITS
    lessly
    0.84
    lessness
    0.83
     chiefly
    0.70
     enormously
    0.70
    pace
    0.70
     tremendously
    0.69
    enance
    0.68
     aloud
    0.67
    iest
    0.66
    ingly
    0.66
    Activations Density 0.021%

    No Known Activations