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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 11-RES-JB
    6. 9682
    Prev
    Next
    INDEX
    Explanations

    names containing "al-" or "-al"

    oai_token-act-pair · gpt-3.5-turbo

    names and titles associated with individuals, particularly those with the prefix "al-"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     revolving
    -0.74
     expanding
    -0.68
     astronauts
    -0.68
     enclosed
    -0.67
     sustained
    -0.66
     investigating
    -0.66
     induction
    -0.65
     encountering
    -0.64
     expansion
    -0.63
     accounting
    -0.63
    POSITIVE LOGITS
    Assad
    1.41
    Islam
    1.41
    Mas
    1.38
    Shar
    1.35
    Kh
    1.34
    Gh
    1.33
    Iss
    1.32
    Sad
    1.30
    Ku
    1.29
    Aw
    1.28
    Activations Density 0.027%

    No Known Activations