© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 751
    Prev
    Next
    INDEX
    Explanations

    mentions of different quantities of items or actions, such as "meaningful actions" or "two distinct tracks"

    oai_token-act-pair · gpt-3.5-turbo

    specific adjectives and descriptors indicating importance or magnitude

    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
    someone
    -0.92
    urgy
    -0.83
    antry
    -0.82
    ylum
    -0.78
    keleton
    -0.74
    rology
    -0.74
    affer
    -0.73
    uddin
    -0.72
    expression
    -0.72
    brance
    -0.70
    POSITIVE LOGITS
     halves
    1.08
     generations
    1.05
     paragraphs
    1.04
     categories
    1.03
     continents
    1.02
     tiers
    1.00
     fronts
    1.00
     thirds
    1.00
     phases
    0.99
     editions
    0.99
    Activations Density 0.210%

    No Known Activations