INDEX

Explanations

names of people or places

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

iasis

-0.71

 discharge

-0.70

 contraceptive

-0.68

staking

-0.67

ramid

-0.64

akening

-0.64

 Piercing

-0.64

 calming

-0.62

aepernick

-0.61

ickr

-0.61

POSITIVE LOGITS

Hug

1.13

glers

0.97

Hug

0.93

 Hugo

0.88

 Pupp

0.85

hett

0.82

Hunt

0.81

gements

0.81

eric

0.80

erd

0.78

Activations Density 0.016%

names of people or places

mentions of specific names or entities

No Comments

No Known Activations