INDEX

Explanations

keywords related to instructions or commands

oai_token-act-pair · gpt-3.5-turbo

the word "For" in various contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 forg

-0.65

 crawl

-0.65

zona

-0.64

itiz

-0.62

soDeliveryDate

-0.58

âĢ³

-0.56

æĺ¯

-0.56

 bottleneck

-0.55

 jaws

-0.55

 helicop

-0.54

POSITIVE LOGITS

bidden

1.62

gotten

1.61

cing

1.21

ced

1.21

give

1.12

 example

1.07

getting

1.04

wards

1.03

 instance

1.01

bid

1.01

Activations Density 0.057%

keywords related to instructions or commands

the word "For" in various contexts

No Comments

No Known Activations