INDEX

Explanations

phrases related to people conversing or discussing certain topics

oai_token-act-pair · gpt-3.5-turbo

phrases or clauses that signal an ongoing or strong action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

utsche

-0.71

iple

-0.68

conom

-0.64

nce

-0.63

©¶æ

-0.63

rius

-0.62

utenant

-0.62

uckland

-0.62

rup

-0.61

chin

-0.61

POSITIVE LOGITS

 albeit

1.31

 culminating

1.07

 prompting

0.94

 namely

0.93

 lest

0.91

etheless

0.87

 regardless

0.86

 includ

0.84

 suggesting

0.84

albeit

0.83

Activations Density 0.586%

phrases related to people conversing or discussing certain topics

phrases or clauses that signal an ongoing or strong action

No Comments

No Known Activations