INDEX

Explanations

terms related to the Environmental Protection Agency (EPA)

oai_token-act-pair · gpt-3.5-turbo

mentions of the Environmental Protection Agency (EPA)

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 latt

-0.68

rave

-0.68

ound

-0.67

 romantic

-0.66

 webs

-0.66

 bitters

-0.65

 simul

-0.62

 dread

-0.61

omb

-0.60

ambo

-0.60

POSITIVE LOGITS

EPA

4.03

EPA

3.23

 Environmental

1.65

CDC

1.55

DOE

1.55

FDA

1.54

 NOAA

1.53

FAA

1.48

 FEMA

1.46

 USDA

1.46

Activations Density 0.010%

terms related to the Environmental Protection Agency (EPA)

mentions of the Environmental Protection Agency (EPA)

No Comments

No Known Activations