INDEX

Explanations

specific words related to the military and locations such as "Pike" and "Shot."

oai_token-act-pair · gpt-3.5-turbo

references to specific products or brands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

enegger

-0.82

schild

-0.81

enance

-0.76

binding

-0.71

 crore

-0.70

orage

-0.69

NUM

-0.68

ieu

-0.68

naire

-0.67

IFIED

-0.67

POSITIVE LOGITS

oples

1.18

ggy

1.09

pe

0.96

cies

0.93

anut

0.91

levard

0.90

achy

0.88

aky

0.88

utics

0.87

ck

0.87

Activations Density 0.011%

specific words related to the military and locations such as "Pike" and "Shot."

references to specific products or brands

No Comments

No Known Activations