INDEX

Explanations

mentions of the word "shares" followed by a numeric value

oai_token-act-pair · gpt-3.5-turbo

references to company shares

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Nost

-0.65

mph

-0.65

pmwiki

-0.65

ALLY

-0.63

 cumbers

-0.63

 Nept

-0.63

 immersion

-0.61

 Manufact

-0.59

IRE

-0.59

ENE

-0.59

POSITIVE LOGITS

hare

1.11

holders

1.06

holder

1.04

 shares

1.03

 Shares

0.98

cro

0.91

Shares

0.86

emonium

0.86

hops

0.83

 share

0.79

Activations Density 0.008%

mentions of the word "shares" followed by a numeric value

references to company shares

No Comments

No Known Activations