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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 6465
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and formatting symbols

    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
    253
    +0.21
    1.5%
    151
    +0.17
    1.2%
    201
    +0.16
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    745
    +0.21
    -0.01
    640
    +0.17
    0.06
    284
    +0.16
    0.06
    Negative Logits
    actly
    -0.79
    hend
    -0.78
    aken
    -0.78
    armac
    -0.77
    auld
    -0.76
     glim
    -0.76
    owship
    -0.76
    isol
    -0.76
    rade
    -0.75
    icult
    -0.75
    POSITIVE LOGITS
    */(
    0.87
     Barclays
    0.81
    ItemTracker
    0.81
     Redemption
    0.74
     Bane
    0.74
    querque
    0.72
    lycer
    0.70
     Jays
    0.70
     à¤
    0.68
     Padres
    0.68
    Activations Density 0.384%

    No Known Activations