Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 6-RES_POST_32K-OAI
    6. 4262
    Prev
    Next
    INDEX
    Explanations

    references to titles, particularly in relation to songs, movies, or characters

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 6-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_6.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.6.hook_resid_post
    Hook Layer
    6
    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
    128
    +0.11
    0.5%
    202
    +0.09
    0.4%
    192
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    651
    +0.11
    0.01
    192
    +0.09
    0.01
    202
    +0.09
    0.01
    Negative Logits
    Downloadha
    -0.70
    hammad
    -0.70
    ijk
    -0.69
    gow
    -0.67
     Sheen
    -0.64
    ¬¼
    -0.63
    agher
    -0.62
     SIG
    -0.61
     Lumpur
    -0.61
     Dhabi
    -0.61
    POSITIVE LOGITS
    anic
    1.05
    ular
    0.93
    entially
    0.93
    ibilities
    0.87
    usha
    0.84
    itans
    0.82
    onic
    0.82
    iton
    0.81
    ania
    0.81
    ented
    0.79
    Activations Density 0.024%

    No Known Activations