INDEX

Explanations

instances of the word "to" followed by verbs, locations, or pronouns

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "return to" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

merce

-0.87

*/(

-0.74

anooga

-0.70

Secondly

-0.69

neum

-0.67

ements

-0.67

nell

-0.66

erity

-0.66

Frameworks

-0.66

Interstitial

-0.65

POSITIVE LOGITS

 haunt

1.14

 basics

1.11

 normal

1.07

 sender

0.90

normal

0.90

 whence

0.90

 square

0.88

 earth

0.85

 prominence

0.85

 norm

0.81

Activations Density 0.087%

instances of the word "to" followed by verbs, locations, or pronouns

instances of the phrase "return to" and its variations

No Comments

No Known Activations