INDEX

Explanations

names and proper nouns related to individuals

oai_token-act-pair · gpt-3.5-turbo

the term "rand" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Referred

-0.79

 thirsty

-0.70

 substance

-0.67

IBLE

-0.67

 towed

-0.66

DMV

-0.65

 Volunteer

-0.65

kens

-0.63

cussion

-0.62

 wastewater

-0.61

POSITIVE LOGITS

rand

1.55

 rand

1.12

olph

1.10

DragonMagazine

1.03

emonium

0.85

uin

0.85

ray

0.84

stad

0.81

zsche

0.80

Rand

0.80

Activations Density 0.010%

names and proper nouns related to individuals

the term "rand" in various contexts

No Comments

No Known Activations