INDEX

Explanations

phrases related to negative judgments on health and fitness

oai_token-act-pair · gpt-3.5-turbo

references to negative health conditions and characteristics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

soType

-0.95

glas

-0.84

lance

-0.81

ership

-0.76

sheets

-0.75

DragonMagazine

-0.74

itivity

-0.74

ibel

-0.74

anmar

-0.73

inical

-0.72

POSITIVE LOGITS

 unnatural

0.80

 wastes

0.76

 costly

0.73

 unreasonable

0.72

 aber

0.71

 contradictory

0.68

 unnecessary

0.67

 unreliable

0.67

bog

0.67

 inefficient

0.66

Activations Density 0.027%

phrases related to negative judgments on health and fitness

references to negative health conditions and characteristics

No Comments

No Known Activations