INDEX

Explanations

instances where something has been seen before or comparisons to prior experiences

oai_token-act-pair · gpt-3.5-turbo

phrases expressing the concept of experiencing or witnessing something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

nai

-0.68

nels

-0.67

apters

-0.64

nel

-0.61

Newsletter

-0.57

emies

-0.57

 reluctantly

-0.56

adium

-0.55

heter

-0.55

 nonetheless

-0.55

POSITIVE LOGITS

 anything

1.17

 anybody

0.97

anything

0.96

 anyone

0.90

ANY

0.82

 Anything

0.76

 such

0.75

him

0.75

nor

0.75

any

0.71

Activations Density 0.088%

instances where something has been seen before or comparisons to prior experiences

phrases expressing the concept of experiencing or witnessing something

No Comments

No Known Activations