INDEX

Explanations

instructions or suggestions for taking specific actions

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing the necessity of verifying or consulting information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

gery

-0.73

hum

-0.71

alf

-0.70

rock

-0.69

bled

-0.68

folk

-0.67

bern

-0.67

MpServer

-0.66

pher

-0.66

OT

-0.66

POSITIVE LOGITS

icio

0.78

 Thrones

0.70

 patience

0.67

 compr

0.67

 beforehand

0.65

 Titus

0.64

ilus

0.62

ppo

0.62

 clicking

0.62

 Siren

0.61

Activations Density 0.037%

instructions or suggestions for taking specific actions

phrases emphasizing the necessity of verifying or consulting information

No Comments

No Known Activations