INDEX

Explanations

mentions of specific software and technology brands, especially Adobe and related products

oai_token-act-pair · gpt-3.5-turbo

references to software and technology brands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ften

-0.81

hillary

-0.77

oldemort

-0.76

ÄŁ

-0.76

ummer

-0.76

0000000000000000

-0.74

sectional

-0.73

nil

-0.72

wright

-0.72

*/(

-0.71

POSITIVE LOGITS

Â®

1.00

 Authent

0.96

 Adapt

0.93

 Certified

0.91

 Remote

0.90

 Flex

0.87

XD

0.87

 Shield

0.86

 Smart

0.86

 Extended

0.85

Activations Density 0.118%

mentions of specific software and technology brands, especially Adobe and related products

references to software and technology brands

No Comments

No Known Activations