INDEX

Explanations

phrases indicating a comparison or contrast

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize the word "so" as a modifier to convey varying degrees of emphasis or comparison

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Aren

-0.64

neum

-0.63

MAP

-0.62

 Ethics

-0.60

(>

-0.59

osterone

-0.58

 Encyclopedia

-0.57

witz

-0.57

IED

-0.56

ertodd

-0.55

POSITIVE LOGITS

 much

1.11

 lucky

0.97

 forgiving

0.94

 subtly

0.91

 easy

0.90

 fortunate

0.88

 simple

0.88

bered

0.87

 easily

0.86

bad

0.85

Activations Density 0.035%

phrases indicating a comparison or contrast

phrases that emphasize the word "so" as a modifier to convey varying degrees of emphasis or comparison

No Comments

No Known Activations