INDEX

Explanations

words related to companies and businesses

oai_token-act-pair · gpt-3.5-turbo

references to significant companies and entities

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

aund

-0.82

 sled

-0.72

obyl

-0.71

izoph

-0.70

veyard

-0.69

Ô

-0.67

pmwiki

-0.67

mire

-0.67

perate

-0.65

kell

-0.65

POSITIVE LOGITS

 headquartered

0.90

 corporation

0.85

 Anthem

0.81

 Siem

0.76

Sony

0.76

ITS

0.74

IG

0.72

 subsidiaries

0.70

SOC

0.70

BM

0.69

Activations Density 0.242%

words related to companies and businesses

references to significant companies and entities

No Comments

No Known Activations