Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 1243
    Prev
    Next
    INDEX
    Explanations

    videos and cameras

    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
    571
    +0.10
    0.4%
    540
    +0.09
    0.4%
    516
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    690
    +0.10
    0.05
    346
    +0.09
    0.06
    540
    +0.09
    0.07
    Negative Logits
    å§«
    -0.82
    Ĭ±
    -0.75
    owder
    -0.70
    cffff
    -0.70
    uckle
    -0.69
    affles
    -0.66
    merce
    -0.66
    vier
    -0.66
     Schwarz
    -0.66
    izabeth
    -0.64
    POSITIVE LOGITS
     footage
    1.24
     clip
    1.22
     clips
    1.21
    clip
    1.03
     filmed
    1.03
     recordings
    0.94
     playback
    0.92
     cameras
    0.91
     uploaded
    0.91
     replay
    0.89
    Activations Density 1.135%

    No Known Activations