INDEX

Explanations

the word "be"

oai_token-act-pair · gpt-3.5-turbo

the occurrence of the verb "be" in different contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nap

-0.77

 lifetime

-0.71

 choice

-0.71

 profile

-0.69

 drift

-0.68

 runaway

-0.67

 relativity

-0.66

 feasibility

-0.65

 firsthand

-0.65

 pole

-0.65

POSITIVE LOGITS

yond

1.43

ardless

1.13

be

1.08

arers

1.07

cker

0.99

efe

0.99

ards

0.99

ech

0.98

FORE

0.96

auc

0.96

Activations Density 0.010%

the word "be"

the occurrence of the verb "be" in different contexts

No Comments

No Known Activations