INDEX

Explanations

adverbs or adjectives that express agreement or approval

oai_token-act-pair · gpt-3.5-turbo

words that indicate correctness or justification

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

stories

-0.70

iments

-0.69

pmwiki

-0.67

 Coffee

-0.65

isms

-0.65

quickShipAvailable

-0.64

 dolls

-0.64

 window

-0.64

facade

-0.62

ession

-0.62

POSITIVE LOGITS

 rightly

1.18

ãĤ©

1.13

 rightfully

1.12

é¾į

0.90

 deserved

0.86

\\\\\\\\\\\\\\\\

0.75

 unfocusedRange

0.75

è¯

0.74

 deserves

0.73

 enthusi

0.72

Activations Density 0.008%

adverbs or adjectives that express agreement or approval

words that indicate correctness or justification

No Comments

No Known Activations