INDEX

Explanations

words related to tasks or responsibilities

oai_token-act-pair · gpt-3.5-turbo

references to assignments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

rians

-0.68

 Britons

-0.66

 hollow

-0.65

BB

-0.64

laugh

-0.64

ooming

-0.64

 Hide

-0.63

oe

-0.63

cept

-0.63

Hyd

-0.62

POSITIVE LOGITS

 assignment

4.05

 assignments

3.31

 Assignment

2.72

 assign

1.81

 assigning

1.68

 assigned

1.66

 assigns

1.52

 allocation

1.37

 homework

1.16

 responsibilities

1.14

Activations Density 0.019%

words related to tasks or responsibilities

references to assignments

No Comments

No Known Activations