INDEX

Explanations

mentions of specific proper nouns or names in various contexts

oai_token-act-pair · gpt-3.5-turbo

words that end in 'er'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

-0.83

ĸļ

-0.66

ournal

-0.62

VIDIA

-0.60

eneg

-0.59

ertodd

-0.59

urer

-0.59

-0.58

 Lovecraft

-0.57

urities

-0.57

POSITIVE LOGITS

jee

1.12

idge

1.02

adish

0.90

unning

0.88

lein

0.85

oute

0.84

aton

0.84

rett

0.82

rors

0.81

wald

0.80

Activations Density 0.088%

mentions of specific proper nouns or names in various contexts

words that end in 'er'

No Comments

No Known Activations