© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 32k
    5. 3-ATT_32K-OAI
    6. 3870
    Prev
    Next
    INDEX
    Explanations

    references to the structure and elements of a narrative plot

    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
    148
    +0.12
    0.6%
    133
    +0.11
    0.5%
    169
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    698
    +0.12
    0.03
    111
    +0.11
    0.03
    427
    +0.10
    0.03
    Negative Logits
    ISA
    -0.79
     nursing
    -0.78
    akens
    -0.71
    íķ
    -0.70
    veland
    -0.70
    ovember
    -0.67
    SPA
    -0.65
     nurs
    -0.65
    ceans
    -0.64
     prized
    -0.63
    POSITIVE LOGITS
     Emin
    0.94
     Vaugh
    0.82
     Circus
    0.79
    plot
    0.75
     Ort
    0.74
    fiction
    0.72
     sqor
    0.72
     unfold
    0.71
     shenan
    0.71
    raq
    0.70
    Activations Density 0.677%

    No Known Activations