INDEX

Explanations

references to different entities related to various locations, such as colleges and countries

oai_token-act-pair · gpt-3.5-turbo

capitalized keywords likely related to specific organizations, locations, or titles

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

idays

-0.82

 srfAttach

-0.69

thodox

-0.66

learn

-0.66

stood

-0.64

hook

-0.64

 overdue

-0.63

repre

-0.61

 unbeliev

-0.61

yip

-0.59

POSITIVE LOGITS

 Rasm

0.72

neck

0.69

estine

0.69

 Phillip

0.67

Daw

0.67

 Ramos

0.66

adelphia

0.66

ordes

0.64

houn

0.64

 Morris

0.63

Activations Density 0.102%

references to different entities related to various locations, such as colleges and countries

capitalized keywords likely related to specific organizations, locations, or titles

No Comments

No Known Activations