INDEX

Explanations

phrases involving step-by-step instructions

oai_token-act-pair · gpt-3.5-turbo

phrases related to sequential instructions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ailability

-0.72

 awakened

-0.63

 Unleashed

-0.62

ortunately

-0.60

 streng

-0.60

ß

-0.59

Pug

-0.59

Hew

-0.59

 notor

-0.58

 Nicotine

-0.58

POSITIVE LOGITS

product

0.99

step

0.91

side

0.90

ward

0.80

example

0.79

products

0.79

slice

0.79

proxy

0.78

case

0.77

lane

0.76

Activations Density 0.030%

phrases involving step-by-step instructions

phrases related to sequential instructions or processes

No Comments

No Known Activations