INDEX

Explanations

ordinal numbers

oai_token-act-pair · gpt-3.5-turbo

references to rankings or placements in comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

uria

-0.64

rina

-0.64

afety

-0.59

uristic

-0.58

teness

-0.58

 Machina

-0.58

arah

-0.56

una

-0.55

ogenesis

-0.55

zar

-0.54

POSITIVE LOGITS

 respectively

2.39

 alike

1.68

 together

1.14

 respective

1.12

 jointly

1.08

together

1.06

both

1.02

depending

0.95

 interchange

0.92

 combined

0.90

Activations Density 0.821%

ordinal numbers

references to rankings or placements in comparisons

No Comments

No Known Activations