© 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. 5438
    Prev
    Next
    INDEX
    Explanations

    numerically quantified changes or conditions in a given context

    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
    135
    +0.12
    0.5%
    141
    +0.10
    0.4%
    741
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.12
    0.01
    714
    +0.10
    0.01
    581
    +0.09
    0.01
    Negative Logits
    utenberg
    -0.66
     acknow
    -0.64
     ·
    -0.58
     misunder
    -0.58
    ccording
    -0.57
     WORLD
    -0.57
     curses
    -0.56
     univers
    -0.56
    tsky
    -0.56
     unab
    -0.55
    POSITIVE LOGITS
    arse
    0.76
    IENT
    0.70
    ergy
    0.69
    auder
    0.66
    rait
    0.66
    hip
    0.64
    maker
    0.62
    lap
    0.62
    ecided
    0.62
    ahime
    0.62
    Activations Density 0.159%

    No Known Activations