INDEX

Explanations

phrases related to learning lessons and acquiring knowledge

oai_token-act-pair · gpt-3.5-turbo

references to learning and gained knowledge or lessons

oai_token-act-pair · gpt-4o-mini

Triggered 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

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

lished

-0.71

vertisement

-0.69

ocument

-0.68

enance

-0.63

cohol

-0.61

 occupancy

-0.61

IUM

-0.61

pite

-0.60

lington

-0.60

istent

-0.59

POSITIVE LOGITS

 lesson

1.46

 lessons

1.23

 Lessons

1.00

 tricks

0.96

 firsthand

0.94

 Learned

0.86

 ropes

0.83

 skills

0.83

 invaluable

0.83

 valuable

0.81

Activations Density 0.199%

phrases related to learning lessons and acquiring knowledge

references to learning and gained knowledge or lessons

No Known Activations