INDEX

Explanations

place names, especially those that may be difficult to pronounce or are related to news events

oai_token-act-pair · gpt-3.5-turbo

references to specific locations and names

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

acy

-0.80

ory

-0.71

uality

-0.69

acea

-0.66

loo

-0.66

angers

-0.62

ORY

-0.60

 calves

-0.60

 saline

-0.60

opian

-0.60

POSITIVE LOGITS

eer

1.07

eering

0.97

lisher

0.95

pload

0.89

eh

0.87

eting

0.86

semb

0.84

fortable

0.84

ebus

0.83

lance

0.79

Activations Density 0.024%

place names, especially those that may be difficult to pronounce or are related to news events

references to specific locations and names

No Comments

No Known Activations