INDEX

Explanations

places or locations, specifically those ending with "ford"

oai_token-act-pair · gpt-3.5-turbo

references to geographical locations or institutions related to "Ford"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ntil

-0.89

Magikarp

-0.81

ngth

-0.77

 exha

-0.71

 federally

-0.71

 tremend

-0.70

ccording

-0.70

Arab

-0.69

 psychiat

-0.69

jriwal

-0.69

POSITIVE LOGITS

shire

1.69

hurst

1.02

ford

0.98

wall

0.95

ness

0.92

leigh

0.86

 Mell

0.83

doms

0.82

vill

0.82

chester

0.81

Activations Density 0.011%

places or locations, specifically those ending with "ford"

references to geographical locations or institutions related to "Ford"

No Comments

No Known Activations