© 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. 5443
    Prev
    Next
    INDEX
    Explanations

    phrases expressing user engagement and activity

    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
    12
    +0.11
    0.5%
    712
    +0.10
    0.5%
    29
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.11
    0.09
    502
    +0.10
    0.07
    49
    +0.09
    0.08
    Negative Logits
     beware
    -0.76
    luck
    -0.70
    liest
    -0.68
     preferred
    -0.68
     slang
    -0.67
     customary
    -0.63
     Concern
    -0.63
     condol
    -0.62
     eyebrows
    -0.61
     Gree
    -0.61
    POSITIVE LOGITS
    ãĥīãĥ©
    0.83
    present
    0.74
    Create
    0.71
    é¾įå
    0.71
     surrounds
    0.70
    create
    0.69
    build
    0.66
     commits
    0.65
     creates
    0.65
     create
    0.65
    Activations Density 0.868%

    No Known Activations