INDEX

Explanations

references to the word "uter" combined with the numbers 10, 9, 8, 7, 6, and 4

oai_token-act-pair · gpt-3.5-turbo

references to regulatory entities or topics related to pollution and environmental concerns

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

ĪĴ

-0.92

ession

-0.83

eling

-0.81

shake

-0.79

paio

-0.78

 challeng

-0.78

eworks

-0.77

elf

-0.76

earchers

-0.75

ness

-0.74

POSITIVE LOGITS

uter

1.08

agonist

0.97

onomy

0.90

onom

0.79

 Parenthood

0.78

 Uran

0.76

itol

0.76

rolet

0.75

ilities

0.75

CHAT

0.75

Activations Density 0.030%

references to the word "uter" combined with the numbers 10, 9, 8, 7, 6, and 4

references to regulatory entities or topics related to pollution and environmental concerns

No Comments

No Known Activations