© 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. 7-RES-JB
    11. 8689
    Prev
    Next
    INDEX
    Explanations

    phrases indicating that a particular action or decision is advantageous or significant

    oai_token-act-pair · gpt-3.5-turbo

    references to actions or events described as "the thing" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ped
    -0.72
     largeDownload
    -0.62
     flats
    -0.61
    illed
    -0.60
    uum
    -0.59
    SPA
    -0.58
     KH
    -0.58
    illing
    -0.57
     sockets
    -0.57
    choes
    -0.57
    POSITIVE LOGITS
     happened
    1.01
     happening
    0.97
     happens
    0.88
     imaginable
    0.88
     happen
    0.85
    Else
    0.77
    OPLE
    0.76
     happ
    0.75
     transpired
    0.71
    iverse
    0.69
    Activations Density 0.031%

    No Known Activations