INDEX

Explanations

words that end with "ret"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "retreat" or related terms hinting at withdrawal or escape

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

CVE

-0.71

 Thames

-0.67

ĪĴ

-0.66

TPPStreamerBot

-0.63

catentry

-0.62

Reviewer

-0.61

ItemThumbnailImage

-0.61

esthesia

-0.61

 GOODMAN

-0.60

xual

-0.60

POSITIVE LOGITS

irement

0.98

rans

0.89

zel

0.87

rent

0.87

ooth

0.84

ail

0.84

rov

0.83

tell

0.83

ret

0.83

rey

0.83

Activations Density 0.013%

words that end with "ret"

variations of the word "retreat" or related terms hinting at withdrawal or escape

No Comments

No Known Activations