© 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. 5412
    Prev
    Next
    INDEX
    Explanations

    constructions of the verb "to be" in various contexts

    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
    320
    +0.11
    0.5%
    373
    +0.10
    0.5%
    20
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.11
    0.09
    566
    +0.10
    0.07
    102
    +0.10
    0.07
    Negative Logits
     miracle
    -0.84
     clich
    -0.78
    successful
    -0.77
     misery
    -0.74
    efficient
    -0.74
     satisfied
    -0.74
     careless
    -0.73
    abuse
    -0.73
     accur
    -0.73
    quit
    -0.72
    POSITIVE LOGITS
     joining
    0.80
    IAL
    0.80
     jointly
    0.77
     livestream
    0.76
     debut
    0.75
     redes
    0.74
     partnering
    0.73
     laun
    0.72
     partnered
    0.71
     unveiling
    0.70
    Activations Density 0.401%

    No Known Activations