INDEX

Explanations

terms related to customer service or protocols like concierge services

oai_token-act-pair · gpt-3.5-turbo

terms related to conciliation and its variants

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

FINEST

-0.72

HY

-0.72

hyde

-0.72

ged

-0.71

 Ames

-0.71

glers

-0.66

doms

-0.64

MENT

-0.61

birds

-0.60

 Unicorn

-0.59

POSITIVE LOGITS

ussions

1.34

urrency

1.31

ussion

1.18

urrent

1.17

urrencies

1.10

ourse

1.06

ourses

1.04

ensus

1.03

urrence

1.03

erto

1.02

Activations Density 0.029%

terms related to customer service or protocols like concierge services

terms related to conciliation and its variants

No Comments

No Known Activations