INDEX

Explanations

questions posing hypothetical scenarios

oai_token-act-pair · gpt-3.5-turbo

questions and phrases directed towards the reader or audience

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

Newsletter

-0.66

MpServer

-0.65

rising

-0.63

 awoke

-0.61

BuyableInstoreAndOnline

-0.60

speaking

-0.60

WithNo

-0.59

 Reborn

-0.59

 awakened

-0.58

 Adin

-0.58

POSITIVE LOGITS

 want

1.36

want

1.17

 prefer

1.11

 wanna

1.08

 prioritize

1.06

 WANT

1.02

 expect

0.98

 choose

0.97

 consider

0.96

 hesitate

0.96

Activations Density 0.268%

questions posing hypothetical scenarios

questions and phrases directed towards the reader or audience

No Comments

No Known Activations