INDEX

Explanations

terms related to political, economic, and industrial subjects

oai_token-act-pair · gpt-3.5-turbo

indicators of uncertainty or doubt

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 favour

-0.64

 invaders

-0.54

 belong

-0.52

 favor

-0.50

Jiu

-0.49

 idol

-0.48

 royalty

-0.48

 roam

-0.47

 whatsoever

-0.47

 chances

-0.47

POSITIVE LOGITS

°

0.84

¾

0.81

·

0.80

¹

0.79

ene

0.79

terness

0.74

enth

0.74

ī

0.73

«

0.73

ctive

0.73

Activations Density 0.559%

terms related to political, economic, and industrial subjects

indicators of uncertainty or doubt

No Comments

No Known Activations