INDEX

Explanations

phrases related to legal matters or authorities taking actions

oai_token-act-pair · gpt-3.5-turbo

mentions of preferential or unfair treatment in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

âĹ¼

-0.91

 tremend

-0.81

 UNCLASSIFIED

-0.76

ngth

-0.71

gom

-0.68

asketball

-0.68

ahime

-0.66

 skelet

-0.65

ESV

-0.63

bowl

-0.63

POSITIVE LOGITS

 frontman

0.71

 watchdog

0.70

 Belfast

0.69

 lawmakers

0.69

 taxpayer

0.68

 lawmaker

0.67

 Donald

0.66

WASHINGTON

0.64

 taxpayers

0.64

 reportedly

0.64

Activations Density 1.331%

phrases related to legal matters or authorities taking actions

mentions of preferential or unfair treatment in various contexts

No Comments

No Known Activations