INDEX

Explanations

terms related to rewards or benefits offered to encourage specific behaviors or actions

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "incentive."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

rooms

-0.91

angers

-0.76

agn

-0.75

âĢ²

-0.73

rup

-0.72

gaard

-0.72

room

-0.70

thing

-0.69

asp

-0.69

afa

-0.67

POSITIVE LOGITS

 incentive

1.30

 incent

1.11

 incentives

1.10

 incentiv

0.97

 compensation

0.90

 compel

0.85

Reviewer

0.85

 compulsion

0.85

 induce

0.84

 rewarded

0.83

Activations Density 0.008%

terms related to rewards or benefits offered to encourage specific behaviors or actions

repeated mentions of the word "incentive."

No Comments

No Known Activations