INDEX

Explanations

mentions of audiences in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to target audiences in content

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

Ide

-0.61

wrong

-0.60

aria

-0.59

 circumstance

-0.58

separ

-0.57

 Heck

-0.57

 polymer

-0.57

 annexation

-0.57

erald

-0.57

Hol

-0.56

POSITIVE LOGITS

 audiences

1.22

 audience

0.94

ÃįÃį

0.89

goers

0.87

hare

0.80

atically

0.79

Reviewer

0.78

fare

0.76

ourcing

0.76

atform

0.75

Activations Density 0.007%

mentions of audiences in various contexts

references to target audiences in content

No Comments

No Known Activations