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

    Neuronpedia

    APICircuit 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. 4-RES-JB
    6. 7696
    Prev
    Next
    INDEX
    Explanations

    instances where the text emphasizes or singles out certain aspects or details

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "especially" indicating emphasis or particularity

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    alities
    -0.73
    lied
    -0.71
    osal
    -0.70
    Runner
    -0.67
    anon
    -0.67
    enary
    -0.66
    hao
    -0.66
    window
    -0.66
    atre
    -0.66
    lying
    -0.65
    POSITIVE LOGITS
     susceptible
    0.86
     noteworthy
    0.78
     influenced
    0.77
     rewarding
    0.75
     suited
    0.74
     exacerbated
    0.73
     advantageous
    0.73
     considering
    0.72
     noticeable
    0.72
     vulnerable
    0.72
    Activations Density 0.026%

    No Known Activations