© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 32k
    5. 3-ATT_32K-OAI
    6. 3379
    Prev
    Next
    INDEX
    Explanations

    instances of the word "discover" and its variations, indicating an interest in discoveries and findings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.3.hook_attn_out
    Hook Layer
    3
    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
    481
    +0.13
    0.6%
    750
    +0.11
    0.5%
    243
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    750
    +0.13
    0.03
    582
    +0.11
    0.03
    576
    +0.11
    0.02
    Negative Logits
    urers
    -0.70
    ince
    -0.65
     farewell
    -0.63
    imet
    -0.62
     upkeep
    -0.60
     assessment
    -0.60
     appe
    -0.60
    onga
    -0.59
    orus
    -0.59
    holders
    -0.59
    POSITIVE LOGITS
    BOOK
    0.72
     caches
    0.71
    ocamp
    0.70
     Applied
    0.68
     similarities
    0.66
     Pengu
    0.66
    ById
    0.66
    furt
    0.66
    0100
    0.65
    Advertisements
    0.65
    Activations Density 0.251%

    No Known Activations