© 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. 5336
    Prev
    Next
    INDEX
    Explanations

    variations of the verb "do" as well as phrases expressing capability or action

    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
    7
    +0.11
    0.5%
    261
    +0.11
    0.5%
    166
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    70
    +0.11
    0.10
    665
    +0.11
    0.09
    761
    +0.10
    0.07
    Negative Logits
    ĸļ
    -0.69
    BuyableInstoreAndOnline
    -0.65
    assault
    -0.64
    IL
    -0.64
    govtrack
    -0.62
    ilty
    -0.61
    mes
    -0.60
    Lenin
    -0.58
    ributes
    -0.55
     meant
    -0.54
    POSITIVE LOGITS
     safely
    1.02
     easily
    1.01
     freely
    0.85
    ibly
    0.83
     cheaply
    0.82
     comfortably
    0.81
    apesh
    0.79
     reliably
    0.76
     indefinitely
    0.74
     afford
    0.73
    Activations Density 0.998%

    No Known Activations