© 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. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4929
    Prev
    Next
    INDEX
    Explanations

    symbols and formatting components within text

    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
    445
    +0.12
    0.6%
    280
    +0.10
    0.5%
    229
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    308
    +0.12
    0.05
    742
    +0.10
    0.07
    280
    +0.10
    0.08
    Negative Logits
    etus
    -0.77
     Ich
    -0.76
    aping
    -0.70
    ophen
    -0.69
     appendix
    -0.69
    emb
    -0.67
    UX
    -0.66
    atl
    -0.66
    IDE
    -0.66
    ancies
    -0.65
    POSITIVE LOGITS
    R
    1.24
     Ri
    1.15
    RP
    1.12
     R
    1.10
    RT
    1.09
    r
    1.08
     r
    1.07
    RG
    1.07
    RR
    1.07
     RD
    1.06
    Activations Density 7.120%

    No Known Activations