© 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. 4739
    Prev
    Next
    INDEX
    Explanations

    instances of punctuation, particularly periods and dashes

    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
    178
    +0.13
    0.6%
    497
    +0.12
    0.6%
    650
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.13
    -0.01
    466
    +0.12
    0.01
    140
    +0.11
    0.01
    Negative Logits
    phony
    -0.68
    Reviewer
    -0.63
     taxp
    -0.59
    SPONSORED
    -0.59
    ndra
    -0.58
     Journals
    -0.56
     menstru
    -0.56
     tsun
    -0.55
     dictators
    -0.53
     Zeit
    -0.52
    POSITIVE LOGITS
    ouver
    0.71
    atro
    0.69
    orio
    0.66
    ividual
    0.66
    acas
    0.65
    lehem
    0.64
    oux
    0.62
    nesota
    0.61
    eele
    0.60
    unicip
    0.60
    Activations Density 0.051%

    No Known Activations