INDEX

Explanations

phrases related to pushing boundaries or exceeding limits

oai_token-act-pair · gpt-3.5-turbo

references to envelopes and related concepts

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

âĵĺ

-0.67

 Spears

-0.65

SOURCE

-0.64

 Haram

-0.63

 WATCHED

-0.63

bot

-0.63

Ķ

-0.63

 Capital

-0.61

ç¥ŀ

-0.61

 acqu

-0.61

POSITIVE LOGITS

 envelope

1.19

velop

1.15

 envelop

1.13

mble

1.05

nuts

0.88

boxes

0.87

eers

0.83

stakes

0.83

leanor

0.81

arding

0.77

Activations Density 0.014%

phrases related to pushing boundaries or exceeding limits

references to envelopes and related concepts

No Comments

No Known Activations