© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4946
    Prev
    Next
    INDEX
    Explanations

    articles and adjectives in the context of categorizing or describing subjects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    568
    +0.11
    0.5%
    218
    +0.10
    0.5%
    564
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.11
    0.14
    362
    +0.10
    0.07
    159
    +0.10
    0.05
    Negative Logits
     farther
    -0.69
     assortment
    -0.69
     distance
    -0.66
     succession
    -0.65
     Sinai
    -0.65
     occas
    -0.63
     vein
    -0.62
     flurry
    -0.61
     slew
    -0.61
     mix
    -0.60
    POSITIVE LOGITS
    successful
    1.08
    ils
    0.91
    gamer
    0.85
    orah
    0.78
    ful
    0.76
    rat
    0.76
    amous
    0.73
    root
    0.72
    ustainable
    0.72
    construct
    0.72
    Activations Density 0.423%

    No Known Activations