INDEX

Explanations

references to geographical locations

oai_token-act-pair · gpt-3.5-turbo

references to the term "local" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

haar

-0.97

xual

-0.92

hower

-0.92

issance

-0.89

ERSON

-0.82

_-

-0.77

lain

-0.76

vous

-0.75

OHN

-0.75

ÙĴ

-0.74

POSITIVE LOGITS

ities

1.02

ised

1.00

isation

0.84

izations

0.83

 grocer

0.83

ization

0.83

izable

0.82

ized

0.82

idad

0.77

 residents

0.77

Activations Density 0.029%

references to geographical locations

references to the term "local" in various contexts

No Comments

No Known Activations