INDEX

Explanations

statements where a definitive answer is provided

oai_token-act-pair · gpt-3.5-turbo

statements emphasizing definitive conclusions or answers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

eatures

-0.71

rongh

-0.70

schild

-0.64

icipated

-0.62

 capacities

-0.62

vre

-0.61

allowed

-0.60

ivities

-0.60

astical

-0.60

 activities

-0.59

POSITIVE LOGITS

YES

1.14

yes

1.14

YES

0.88

 unequiv

0.87

nil

0.87

yes

0.85

 affirmative

0.83

NO

0.81

 Nope

0.79

 nowhere

0.78

Activations Density 0.101%

statements where a definitive answer is provided

statements emphasizing definitive conclusions or answers

No Comments

No Known Activations