© 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. 0-RES-JB
    11. 20434
    Prev
    Next
    INDEX
    Explanations

    words related to a specific body part or an area of the body

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "hushed" or variations associated with hushed or quiet environments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Obst
    -0.69
    onge
    -0.67
    esthesia
    -0.66
    hex
    -0.66
    nect
    -0.65
    emort
    -0.65
    ifest
    -0.64
    usterity
    -0.64
     anatomy
    -0.64
    hemer
    -0.64
    POSITIVE LOGITS
    è¦ļéĨĴ
    0.85
    Book
    0.76
     Listen
    0.69
    Phoenix
    0.68
    mir
    0.68
    Assistant
    0.67
    senal
    0.65
     supers
    0.64
    Fourth
    0.63
    TON
    0.62
    Activations Density 0.000%

    No Known Activations