© 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. 19949
    Prev
    Next
    INDEX
    Explanations

    words related to actions or objects that have been captured or secured

    oai_token-act-pair · gpt-3.5-turbo

    words related to various forms of injury, ailments, or conditions

    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
    hill
    -0.65
     imperson
    -0.59
     Leg
    -0.56
     tally
    -0.55
     clock
    -0.55
    fax
    -0.55
     interference
    -0.54
     specifically
    -0.54
    bos
    -0.54
     Bulletin
    -0.53
    POSITIVE LOGITS
    ured
    4.54
    uring
    2.87
    ures
    2.86
    urers
    2.33
    ure
    2.32
    urer
    2.14
    urable
    2.08
    URES
    2.04
    URE
    2.03
    urally
    1.91
    Activations Density 0.007%

    No Known Activations