INDEX

Explanations

mentions of the FDA

oai_token-act-pair · gpt-3.5-turbo

references to the FDA

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

atre

-0.70

john

-0.68

Else

-0.67

 Poles

-0.67

 Sleeping

-0.67

Riy

-0.66

semble

-0.65

 Celtic

-0.65

lihood

-0.64

 Syri

-0.63

POSITIVE LOGITS

 filing

0.74

 approves

0.74

 bureaucracy

0.74

FDA

0.73

 approved

0.72

 approve

0.71

 regulators

0.71

 regulates

0.71

 checkpoint

0.70

 councill

0.70

Activations Density 0.014%

mentions of the FDA

references to the FDA

No Comments

No Known Activations