INDEX

Explanations

dates from the mid-20th century

oai_token-act-pair · gpt-3.5-turbo

references to significant historical years

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

parency

-0.71

edo

-0.69

resso

-0.68

clus

-0.65

venge

-0.65

opher

-0.65

hed

-0.64

 verbs

-0.63

unity

-0.62

 impunity

-0.61

POSITIVE LOGITS

å¹

0.77

-'

0.75

keley

0.73

çļ

0.72

0.65

 MacArthur

0.64

 Reincarn

0.64

0.63

 Reich

0.63

Activations Density 0.041%

dates from the mid-20th century

references to significant historical years

No Comments

No Known Activations