INDEX

Explanations

words related to subsequent actions or events, specifically words related to continuing or following up on something

oai_token-act-pair · gpt-3.5-turbo

phrases related to follow-up actions or events

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

 privately

-0.71

 pulp

-0.66

yip

-0.63

 constitu

-0.62

 rusty

-0.61

 opener

-0.61

 soluble

-0.60

 nerv

-0.59

 interchange

-0.58

job

-0.57

POSITIVE LOGITS

advertisement

1.00

release

0.96

up

0.91

beat

0.90

appro

0.87

established

0.86

cart

0.85

growth

0.84

action

0.84

recomm

0.84

Activations Density 0.043%

words related to subsequent actions or events, specifically words related to continuing or following up on something

phrases related to follow-up actions or events

No Comments

No Known Activations