INDEX

Explanations

references to "Warner Bros" specifically

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of "Bros" in various contexts, likely reflecting on the Warner Bros studio and its productions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

pod

-0.70

 Soviets

-0.67

 Soviet

-0.64

 FEMA

-0.63

Soviet

-0.62

 Austrian

-0.61

Reviewer

-0.61

CBO

-0.60

PLIED

-0.59

 substitute

-0.59

POSITIVE LOGITS

 Bros

1.03

Ã»

0.86

ingen

0.85

hes

0.81

orp

0.80

uits

0.78

ke

0.77

anes

0.76

gen

0.76

olt

0.76

Activations Density 0.007%

references to "Warner Bros" specifically

repeated mentions of "Bros" in various contexts, likely reflecting on the Warner Bros studio and its productions

No Comments

No Known Activations