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. 3-RES-JB
    6. 1012
    Prev
    Next
    INDEX
    Explanations

    phrases containing the word "naught" which often signify a negative outcome or insignificance

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to feelings of futility or lack of value

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     diffusion
    -0.64
     occupancy
    -0.63
     characterize
    -0.61
     limp
    -0.60
     blackout
    -0.60
     dens
    -0.60
     affinity
    -0.59
     annexation
    -0.59
     reaction
    -0.58
     atmosp
    -0.58
    POSITIVE LOGITS
    aught
    1.10
    ead
    0.85
    shire
    0.78
    tackle
    0.77
    erer
    0.76
    framework
    0.76
    ãĤ¤ãĥĪ
    0.76
    rings
    0.75
    ãĤ¦ãĤ¹
    0.75
    nesses
    0.74
    Activations Density 0.012%

    No Known Activations