INDEX

Explanations

words related to gracefully accepting or considering something

oai_token-act-pair · gpt-3.5-turbo

adverbs indicating completeness or thoroughness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Greenwich

-0.67

 wallets

-0.66

Reviewer

-0.66

Frameworks

-0.66

senal

-0.64

 bills

-0.63

 budgets

-0.63

OTOS

-0.63

 extrem

-0.62

soDeliveryDate

-0.61

POSITIVE LOGITS

ctuary

0.89

fully

0.87

theless

0.87

ful

0.84

wash

0.83

xtap

0.81

fulness

0.80

hyd

0.73

nergy

0.72

 regenerate

0.72

Activations Density 0.005%

words related to gracefully accepting or considering something

adverbs indicating completeness or thoroughness

No Comments

No Known Activations