© 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 49152
    9. 8-RES_FS49152-JB
    11. 26554
    Prev
    Next
    INDEX
    Explanations

    terms related to roles and positions of individuals or groups in various contexts

    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
    481
    +0.17
    0.8%
    447
    +0.13
    0.6%
    215
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    537
    +0.17
    0.06
    575
    +0.13
    0.06
    224
    +0.10
    0.05
    Negative Logits
     increments
    -0.70
    yss
    -0.70
    rawdownloadcloneembedreportprint
    -0.64
     pestic
    -0.61
    omial
    -0.60
    Hig
    -0.59
    OUS
    -0.58
     shakes
    -0.58
    LOCK
    -0.58
     acknow
    -0.58
    POSITIVE LOGITS
    hips
    1.22
    hip
    1.21
     alike
    1.14
     who
    1.06
     whom
    1.00
    paces
    0.92
    ervatives
    0.90
     whose
    0.88
    who
    0.87
    '
    0.87
    Activations Density 0.190%

    No Known Activations