INDEX

Explanations

the word "Watson" with various prefixes and suffixes

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Watson."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ccording

-0.80

 fertile

-0.78

cised

-0.78

lav

-0.76

ional

-0.76

ĻĤ

-0.76

ript

-0.74

haps

-0.74

wives

-0.73

eways

-0.73

POSITIVE LOGITS

 Watson

1.54

atson

1.11

 Hawkins

0.96

 Holmes

0.93

Whe

0.90

combe

0.88

bies

0.84

 Jarvis

0.83

 Evans

0.82

 Cumm

0.80

Activations Density 0.004%

the word "Watson" with various prefixes and suffixes

occurrences of the name "Watson."

No Comments

No Known Activations