INDEX

Explanations

mentions of health-related topics, such as medical conditions, behaviors, and therapies

oai_token-act-pair · gpt-3.5-turbo

topics related to negative social issues and consequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 furiously

-0.73

 ener

-0.65

 fierce

-0.64

 beautifully

-0.62

 earnest

-0.61

 tantal

-0.61

 brilliantly

-0.60

 frantic

-0.59

 onlook

-0.59

 brazen

-0.59

POSITIVE LOGITS

 depends

1.11

Â¶

1.08

 varies

1.04

.--

0.98

 =================

0.98

(%)

0.97

 =================================

0.93

 shall

0.93

):

0.93

 determines

0.92

Activations Density 0.770%

mentions of health-related topics, such as medical conditions, behaviors, and therapies

topics related to negative social issues and consequences

No Comments

No Known Activations