INDEX

Explanations

questions being asked

oai_token-act-pair · gpt-3.5-turbo

questions or phrases expressing a desire to acquire information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ItemTracker

-0.94

phrine

-0.80

ŃĶ

-0.77

ĪĴ

-0.76

interstitial

-0.70

©¶æ

-0.69

ovie

-0.69

pite

-0.65

©¶æ¥µ

-0.65

ountain

-0.64

POSITIVE LOGITS

ledge

1.07

lege

0.96

how

0.94

ariat

0.84

afer

0.83

why

0.80

why

0.80

WHY

0.78

HOW

0.77

 ABOUT

0.75

Activations Density 0.048%

questions being asked

questions or phrases expressing a desire to acquire information

No Comments

No Known Activations