INDEX

Explanations

mentions of various types of worms and the word 'Worms' with a capital 'W'

oai_token-act-pair · gpt-3.5-turbo

references to worms and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 arresting

-0.73

owered

-0.72

ourt

-0.71

orney

-0.71

orically

-0.70

 Palestin

-0.70

 seiz

-0.70

ournal

-0.69

liction

-0.69

yles

-0.69

POSITIVE LOGITS

hole

1.36

worms

1.22

holes

1.15

worm

1.13

tail

1.13

roots

1.01

 worms

0.91

fish

0.87

 Worm

0.83

pool

0.82

Activations Density 0.041%

mentions of various types of worms and the word 'Worms' with a capital 'W'

references to worms and related terms

No Comments

No Known Activations