INDEX

Explanations

email-related prompts and instructions

oai_token-act-pair · gpt-3.5-turbo

email-related requests and validations

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

imer

-0.88

iaries

-0.75

mons

-0.68

farm

-0.65

Thumbnail

-0.65

geist

-0.63

aimon

-0.61

igans

-0.61

 Bend

-0.61

leys

-0.60

POSITIVE LOGITS

 Cancel

0.87

$$$$

0.75

ause

0.68

 Schwarzenegger

0.67

{*

0.65

 invalid

0.64

icol

0.64

URL

0.63

ername

0.63

herry

0.62

Activations Density 0.067%

email-related prompts and instructions

email-related requests and validations

No Comments

No Known Activations