INDEX

Explanations

negative statements or denials

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the absence or lack of something

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

parts

-0.96

redients

-0.90

agents

-0.89

ents

-0.89

devices

-0.88

anners

-0.88

items

-0.87

projects

-0.85

 Actions

-0.85

eals

-0.83

POSITIVE LOGITS

 indication

1.41

 shortage

1.28

 doubt

1.15

 guarantee

1.13

 suggestion

1.03

 evidence

1.03

 reason

1.03

way

1.02

 requirement

1.01

 denying

0.99

Activations Density 0.058%

negative statements or denials

phrases indicating the absence or lack of something

No Comments

No Known Activations