INDEX

Explanations

mentions of specific names or locations

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, events, or numbers in a context that suggests a narrative or factual information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

•EggList

No Comments

Negative Logits

Ply

-0.83

 WiFi

-0.69

 stre

-0.68

 Tenth

-0.65

LAR

-0.65

LIN

-0.64

 Stellar

-0.63

LOS

-0.63

 slic

-0.62

 Winston

-0.62

POSITIVE LOGITS

aq

1.12

AC

1.07

amate

1.07

ac

1.06

af

1.05

aed

1.03

abis

1.02

atical

1.02

atar

1.01

ab

1.00

Activations Density 0.463%

mentions of specific names or locations

references to specific individuals, events, or numbers in a context that suggests a narrative or factual information

No Comments

No Known Activations