INDEX

Explanations

proper nouns or names potentially related to locations or people

oai_token-act-pair · gpt-3.5-turbo

references to "Bra" and its variations in various contexts

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

eers

-0.78

omez

-0.75

papers

-0.71

matic

-0.70

ership

-0.70

ancial

-0.67

aires

-0.67

geist

-0.66

aminer

-0.66

regor

-0.65

POSITIVE LOGITS

ided

1.10

ille

1.03

hma

0.99

kes

0.96

very

0.94

zz

0.92

keley

0.92

iding

0.91

gger

0.90

ken

0.87

Activations Density 0.042%

proper nouns or names potentially related to locations or people

references to "Bra" and its variations in various contexts

No Comments

No Known Activations