INDEX

Explanations

verbs that suggest taking action or making progress

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate actions initiated by a subject

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

adal

-0.67

Silver

-0.65

ario

-0.63

ime

-0.62

>)

-0.62

atal

-0.62

atem

-0.60

jon

-0.60

zman

-0.59

atl

-0.59

POSITIVE LOGITS

 virtue

1.27

 placing

1.17

 allowing

1.17

 adding

1.13

 eliminating

1.13

 creating

1.13

 removing

1.11

 providing

1.10

 sacrificing

1.08

products

1.08

Activations Density 0.102%

verbs that suggest taking action or making progress

phrases that indicate actions initiated by a subject

No Comments

No Known Activations