© 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. 1924
    Prev
    Next
    INDEX
    Explanations

    terms related to citations and statements made by individuals

    oai_token-act-pair · gpt-3.5-turbo

    references to authors or individuals making statements or remarks

    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
     perfected
    -0.50
    adies
    -0.48
     assassinate
    -0.47
     performers
    -0.47
     enrich
    -0.45
     matured
    -0.45
     puppies
    -0.45
    ablish
    -0.44
    ardless
    -0.43
     proven
    -0.43
    POSITIVE LOGITS
     spokesman
    0.63
     spokeswoman
    0.60
     Xin
    0.56
    veland
    0.56
    soType
    0.54
     spokesperson
    0.52
     Moff
    0.50
    IPS
    0.49
    aretz
    0.49
     Securities
    0.48
    Activations Density 0.674%

    No Known Activations