© 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. 10-RES-JB
    11. 23435
    Prev
    Next
    INDEX
    Explanations

    terms related to large scale events, such as "biggest", "largest", "deadliest", and "longest"

    oai_token-act-pair · gpt-3.5-turbo

    superlative adjectives and phrases indicating the largest, biggest, or most significant instances of events or items

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    instead
    -0.78
    fen
    -0.75
    kas
    -0.71
    olith
    -0.69
    lishes
    -0.68
    vantage
    -0.68
    abet
    -0.68
     Regist
    -0.67
    acent
    -0.67
    ied
    -0.66
    POSITIVE LOGITS
     ever
    1.40
     EVER
    1.24
    ever
    1.20
     imaginable
    1.15
     peac
    0.96
     Ever
    0.90
     single
    0.83
     uninterrupted
    0.82
     conceivable
    0.81
     possible
    0.78
    Activations Density 0.109%

    No Known Activations