© 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. 1-RES-JB
    11. 13322
    Prev
    Next
    INDEX
    Explanations

    words related to impact or influence on different entities

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "affect" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Ou
    -0.75
    Vs
    -0.72
     Tan
    -0.68
    Walk
    -0.67
    SEC
    -0.66
    YC
    -0.66
    VIDEO
    -0.66
     Vor
    -0.66
    ALK
    -0.66
    ATT
    -0.65
    POSITIVE LOGITS
     affect
    1.13
     adversely
    1.08
    uate
    1.01
    uates
    1.00
     impact
    0.97
     affects
    0.96
    uated
    0.90
     impacts
    0.90
     affected
    0.90
     havoc
    0.87
    Activations Density 0.013%

    No Known Activations