INDEX

Explanations

statements made by individuals in news or interview settings

oai_token-act-pair · gpt-3.5-turbo

statements related to political issues and controversies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 aspirin

-0.68

 storefront

-0.65

 totality

-0.61

cox

-0.60

 lantern

-0.60

 basil

-0.59

 bandits

-0.59

 crimson

-0.58

 moth

-0.58

 goblins

-0.58

POSITIVE LOGITS

Speaking

1.58

 Speaking

1.35

Asked

1.24

Interested

1.21

Writing

1.15

Refer

1.07

Talking

1.06

Discuss

1.00

Calling

0.98

 Asked

0.98

Activations Density 0.395%

statements made by individuals in news or interview settings

statements related to political issues and controversies

No Comments

No Known Activations