INDEX

Explanations

companies and entities mentioned in a document

oai_token-act-pair · gpt-3.5-turbo

references to companies and corporations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

theless

-0.66

rawdownloadcloneembedreportprint

-0.64

shows

-0.64

hus

-0.63

Maw

-0.63

LIB

-0.62

 Enlightenment

-0.62

WARE

-0.61

ength

-0.61

ttp

-0.60

POSITIVE LOGITS

 headquartered

0.94

 executives

0.93

 CEOs

0.87

folk

0.87

afety

0.83

mens

0.82

 companies

0.82

wide

0.81

 nomine

0.81

orpor

0.80

Activations Density 0.044%

companies and entities mentioned in a document

references to companies and corporations

No Comments

No Known Activations