INDEX

Explanations

phrases related to discovering new information or objects

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "discovered."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

quote

-0.70

yip

-0.70

erate

-0.66

oided

-0.66

cussion

-0.65

wrap

-0.65

Ã±a

-0.64

wr

-0.64

quet

-0.63

 paced

-0.62

POSITIVE LOGITS

 Lauder

0.78

Discover

0.76

âĸĴ

0.75

Ô

0.75

çīĪ

0.74

æĥ

0.72

gdala

0.71

ibility

0.71

 flaws

0.71

Frameworks

0.71

Activations Density 0.032%

phrases related to discovering new information or objects

occurrences of the word "discovered."

No Comments

No Known Activations