© 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. 6382
    Prev
    Next
    INDEX
    Explanations

    technical terms related to microscopy and immunostaining procedures

    oai_token-act-pair · gpt-3.5-turbo

    technical terms and phrases related to scientific experiments and processes

    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
    conom
    -0.80
    女
    -0.76
    estate
    -0.74
     perks
    -0.74
     someday
    -0.73
     Stranger
    -0.72
     somet
    -0.71
     unfairly
    -0.71
    ulhu
    -0.70
     tru
    -0.69
    POSITIVE LOGITS
     assay
    1.21
     pH
    1.14
     PCR
    1.13
     α
    1.03
     ÎĶ
    1.03
     purified
    1.02
     cyt
    1.01
     enz
    0.99
     phosph
    0.99
     saline
    0.98
    Activations Density 0.590%

    No Known Activations