INDEX

Explanations

warnings or alerts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "warning."

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

morph

-0.84

dx

-0.80

animate

-0.77

rencies

-0.73

adr

-0.73

anova

-0.73

ophon

-0.72

rafted

-0.72

tiny

-0.72

growth

-0.72

POSITIVE LOGITS

 warning

1.18

 Warning

1.17

 warnings

1.07

 Warn

1.04

warning

1.03

 warn

0.99

 warns

0.95

 caution

0.92

warn

0.91

 disclaimer

0.85

Activations Density 0.018%

warnings or alerts

instances of the word "warning."

No Comments

No Known Activations