© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5720
    Prev
    Next
    INDEX
    Explanations

    references to counter or opposing arguments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    578
    +0.12
    0.6%
    314
    +0.11
    0.5%
    222
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    222
    +0.12
    0.02
    3
    +0.11
    0.02
    578
    +0.10
    0.02
    Negative Logits
     ABE
    -0.70
     Forge
    -0.66
    OOD
    -0.66
    doms
    -0.66
     Awards
    -0.63
     Bram
    -0.60
     Tornado
    -0.59
     throats
    -0.59
     Flesh
    -0.58
    åĬ
    -0.58
    POSITIVE LOGITS
    measures
    1.30
    balance
    1.19
    intuitive
    1.17
    clock
    1.16
    attack
    1.14
    tops
    1.09
    pell
    1.03
    fe
    1.03
    balanced
    1.01
    party
    1.01
    Activations Density 0.051%

    No Known Activations