INDEX

Explanations

phrases related to agreements or deals being made

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases related to agreements or contracts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

ãĥķãĤ¡

-0.77

espie

-0.72

âĢ¢âĢ¢âĢ¢âĢ¢

-0.71

Focus

-0.69

Anonymous

-0.69

Learning

-0.68

 Citiz

-0.68

CENT

-0.66

 Focus

-0.65

Posts

-0.65

POSITIVE LOGITS

 consum

1.18

 renegoti

1.13

vo

1.07

 negotiated

1.06

 signed

1.04

 ratified

1.03

 struck

1.02

 finalized

0.99

 concluded

0.98

 broke

0.95

Activations Density 0.161%

phrases related to agreements or deals being made

phrases related to agreements or contracts

No Comments

No Known Activations