INDEX

Explanations

phrases indicating to monitor or pay attention to something

oai_token-act-pair · gpt-3.5-turbo

references to surveillance or keeping track of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Ì¶

-0.86

©¶æ¥µ

-0.85

hement

-0.74

itle

-0.73

phabet

-0.71

ayers

-0.69

 sadd

-0.69

eno

-0.68

gebra

-0.68

Â¢

-0.67

POSITIVE LOGITS

 closely

1.03

 developments

0.96

 lest

0.90

 suspicious

0.84

 incoming

0.82

 trends

0.81

 fluctuations

0.72

 whereabouts

0.71

 periphery

0.71

 carefully

0.70

Activations Density 0.178%

phrases indicating to monitor or pay attention to something

references to surveillance or keeping track of something

No Comments

No Known Activations