© 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. 3927
    Prev
    Next
    INDEX
    Explanations

    financial and economic terms related to community resources and events

    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
    745
    +0.20
    0.9%
    420
    +0.13
    0.6%
    586
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    57
    +0.20
    0.08
    620
    +0.13
    0.10
    30
    +0.10
    0.10
    Negative Logits
    govtrack
    -0.84
     Dunn
    -0.76
    mediately
    -0.76
    izont
    -0.68
    rily
    -0.67
     abruptly
    -0.67
    usal
    -0.65
     inning
    -0.65
     Coffin
    -0.63
    rils
    -0.62
    POSITIVE LOGITS
     =================================
    0.77
    heid
    0.70
     Attribution
    0.67
     Languages
    0.67
     Navigation
    0.67
     physic
    0.66
     proble
    0.66
    ophon
    0.65
     sugg
    0.65
     Compan
    0.65
    Activations Density 14.531%

    No Known Activations