INDEX

Explanations

references to the word "ign" in varying contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the substring "ign."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

bian

-0.78

HAHAHAHA

-0.72

 Springfield

-0.68

cham

-0.67

------------------------------------------------

-0.67

 Span

-0.65

Bie

-0.65

————————

-0.64

 Gingrich

-0.63

successfully

-0.61

POSITIVE LOGITS

ments

1.18

ificant

1.05

antly

1.05

mentation

0.99

entials

0.98

eous

0.95

ame

0.94

atories

0.90

atures

0.90

iew

0.89

Activations Density 0.013%

references to the word "ign" in varying contexts

instances of the substring "ign."

No Comments

No Known Activations