INDEX

Explanations

terms related to political entities, such as countries, organizations, and regions

oai_token-act-pair · gpt-3.5-turbo

connector words and punctuation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 âĸĪ

-0.45

 Trooper

-0.44

isode

-0.43

 Instruct

-0.43

rary

-0.43

âĸĪ

-0.42

uates

-0.42

hyde

-0.41

fold

-0.41

itial

-0.41

POSITIVE LOGITS

rimp

0.47

 vanquished

0.46

 amphib

0.45

 multinational

0.45

 Caucasus

0.44

 medd

0.44

 sclerosis

0.43

 resurg

0.42

 neighbouring

0.41

 Bangladesh

0.40

Activations Density 1.530%

terms related to political entities, such as countries, organizations, and regions

connector words and punctuation

No Comments

No Known Activations