INDEX

Explanations

phrases related to exploration and discovery

oai_token-act-pair · gpt-3.5-turbo

instances of the word "explore" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

fixed

-0.73

jud

-0.71

die

-0.70

fight

-0.65

wa

-0.65

absolute

-0.64

 disqualified

-0.64

enta

-0.64

death

-0.63

ĪĴ

-0.63

POSITIVE LOGITS

ĸļ

0.97

 Explore

0.85

çīĪ

0.85

ibilities

0.81

ibility

0.80

ationally

0.75

 exploring

0.75

URES

0.74

prising

0.73

ONSORED

0.72

Activations Density 0.019%

phrases related to exploration and discovery

instances of the word "explore" and its variations

No Comments

No Known Activations