INDEX

Explanations

places, events, and people during interviews or speeches

oai_token-act-pair · gpt-3.5-turbo

instances of the word "the" and expressions denoting events or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

strap

-0.83

alties

-0.78

 besides

-0.74

 experimented

-0.72

 owing

-0.71

 survives

-0.70

Layer

-0.68

 someday

-0.68

 somehow

-0.66

 suppose

-0.66

POSITIVE LOGITS

 latest

1.01

 latter

0.95

 announcement

0.93

 inaugural

0.92

 sidelines

0.91

 Huffington

0.89

 podium

0.88

 Kremlin

0.87

 outset

0.86

 nation

0.85

Activations Density 0.430%

places, events, and people during interviews or speeches

instances of the word "the" and expressions denoting events or activities

No Comments

No Known Activations