INDEX

Explanations

phrases indicating a future action or intention

oai_token-act-pair · gpt-3.5-turbo

instances of the word "will" indicating future actions or predictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Newsletter

-0.78

amped

-0.66

eatured

-0.62

existence

-0.61

 Romeo

-0.61

aina

-0.61

efeated

-0.60

cius

-0.60

Methods

-0.60

Pie

-0.60

POSITIVE LOGITS

be

1.08

 gladly

1.06

 doubtless

1.04

 probably

1.03

 surely

1.02

 undoubtedly

1.01

 likely

0.98

 definitely

0.96

 inevitably

0.95

 eventually

0.95

Activations Density 0.219%

phrases indicating a future action or intention

instances of the word "will" indicating future actions or predictions

No Comments

No Known Activations