© 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. 3-RES-JB
    11. 6909
    Prev
    Next
    INDEX
    Explanations

    phrases related to an artwork or creation being described as a "masterpiece"

    oai_token-act-pair · gpt-3.5-turbo

    references to distinct segments or components referred to as 'pieces'

    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
     footh
    -0.75
    pathy
    -0.65
    âĸ¬
    -0.63
     equations
    -0.63
     controllers
    -0.62
     sciences
    -0.61
     AMA
    -0.60
    iour
    -0.60
     burn
    -0.59
     tomat
    -0.59
    POSITIVE LOGITS
    meal
    1.34
    piece
    0.95
    OGR
    0.93
    enary
    0.87
    holder
    0.85
    glass
    0.83
    acement
    0.82
    fare
    0.81
    ngth
    0.81
    ridges
    0.81
    Activations Density 0.015%

    No Known Activations