© 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. 5411
    Prev
    Next
    INDEX
    Explanations

    the article "a" and its frequency in the document

    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
    685
    +0.11
    0.5%
    610
    +0.11
    0.5%
    362
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.11
    0.07
    512
    +0.11
    0.04
    362
    +0.10
    0.04
    Negative Logits
    ulhu
    -0.79
    ezvous
    -0.66
     Nanto
    -0.64
     unfocusedRange
    -0.64
    ulative
    -0.64
    igators
    -0.61
    senal
    -0.61
    olson
    -0.61
     Tid
    -0.60
    berus
    -0.60
    POSITIVE LOGITS
    dozen
    1.16
     dozen
    1.08
     hundred
    1.00
     thousand
    0.93
     decade
    0.83
     century
    0.81
     million
    0.78
     millennium
    0.76
    ĪĴ
    0.74
     lakh
    0.72
    Activations Density 0.169%

    No Known Activations