© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 23782
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a negative impact or difficulty being faced

    oai_token-act-pair · gpt-3.5-turbo

    various contexts where the concepts of "matters," "life," and "things" are discussed, particularly in relation to making situations either better or worse

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arettes
    -0.70
    sha
    -0.68
    owment
    -0.68
    acted
    -0.64
    ennial
    -0.64
    atan
    -0.62
    lav
    -0.62
    ority
    -0.62
    berra
    -0.61
     participated
    -0.60
    POSITIVE LOGITS
     manageable
    0.66
     Puzz
    0.65
     unpop
    0.64
     easier
    0.63
     nicely
    0.63
     simpler
    0.61
     ticking
    0.60
     tighter
    0.58
     smoother
    0.58
     heating
    0.58
    Activations Density 0.108%

    No Known Activations