INDEX

Explanations

emphatic statements conveying absolutes or extremes

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing the concepts of never or always in relation to experiences or events

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

 hire

-0.67

 think

-0.65

 thinks

-0.65

 undertake

-0.65

 expects

-0.64

 spends

-0.63

 watches

-0.62

cule

-0.62

 cares

-0.61

 realize

-0.61

POSITIVE LOGITS

been

1.50

 been

1.47

 occurred

1.42

 arisen

1.33

 Been

1.32

 happened

1.27

 resulted

1.26

 transpired

1.24

 flowed

1.19

 existed

1.15

Activations Density 0.144%

emphatic statements conveying absolutes or extremes

phrases emphasizing the concepts of never or always in relation to experiences or events

No Comments

No Known Activations