© 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. Attention Out - 32k
    9. 3-ATT_32K-OAI
    11. 3981
    Prev
    Next
    INDEX
    Explanations

    phrases related to low-cost or low-voltage concepts

    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
    420
    +0.13
    0.6%
    497
    +0.10
    0.4%
    701
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    132
    +0.13
    0.05
    540
    +0.10
    0.04
    35
    +0.09
    0.03
    Negative Logits
    PLA
    -0.98
    çīĪ
    -0.79
    YES
    -0.78
     OU
    -0.78
    frey
    -0.77
     Including
    -0.76
    alogy
    -0.72
    andise
    -0.71
    ZA
    -0.70
    ozy
    -0.70
    POSITIVE LOGITS
    luster
    0.72
     anymore
    0.71
    ocr
    0.67
     downtime
    0.67
    anova
    0.63
    cy
    0.62
    til
    0.62
     cents
    0.62
    warm
    0.62
     morals
    0.62
    Activations Density 0.443%

    No Known Activations