INDEX

Explanations

numerical values that are 89 or similar patterns

oai_token-act-pair · gpt-3.5-turbo

the repeated mention of the number 89 and the year 1989

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arent

-0.79

 Dull

-0.75

istical

-0.72

åĤ

-0.71

ggies

-0.70

 Nadu

-0.69

ategic

-0.69

ional

-0.68

ugu

-0.66

 Contra

-0.65

POSITIVE LOGITS

0.83

0.81

0.80

0.78

0.76

0.75

ourn

0.74

0.73

Activations Density 0.029%

numerical values that are 89 or similar patterns

the repeated mention of the number 89 and the year 1989

No Comments

No Known Activations