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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 128k
    5. 0-ATT_128K-OAI
    6. 13559
    Prev
    Next
    INDEX
    Explanations

    terms related to ambivalence and atmospheric qualities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/GPT2-Small-OAI-v5-128k-attn-out-SAEs/v5_128k_layer_0
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    131,072
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_attn_out
    Hook Layer
    0
    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
    211
    +0.18
    1.2%
    153
    +0.18
    1.2%
    201
    +0.16
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    153
    +0.18
    0.01
    56
    +0.18
    0.01
    51
    +0.16
    0.01
    Negative Logits
     aggregate
    -0.78
    wcsstore
    -0.75
    Month
    -0.72
    amins
    -0.72
    natureconservancy
    -0.70
    DEV
    -0.69
    dayName
    -0.68
    ULE
    -0.68
    ãĥķãĤ©
    -0.67
    Tumblr
    -0.67
    POSITIVE LOGITS
     amb
    0.89
    doors
    0.78
     seated
    0.77
    otom
    0.74
     voices
    0.70
     Tanz
    0.70
     seiz
    0.70
     robbers
    0.70
    guards
    0.68
    cade
    0.68
    Activations Density 0.052%

    No Known Activations