INDEX

Explanations

instances where a review or consideration is being conducted or considered

oai_token-act-pair · gpt-3.5-turbo

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Notable

-0.72

âĹ¼

-0.71

Buyable

-0.61

 Kinnikuman

-0.59

Hero

-0.59

 bearer

-0.58

breaks

-0.58

 utter

-0.57

 abbre

-0.57

ModLoader

-0.56

POSITIVE LOGITS

 process

1.10

 Panel

0.99

 panel

0.97

 board

0.95

 chaired

0.91

 committee

0.89

 period

0.87

 deliber

0.87

 Process

0.85

 hearings

0.85

Activations Density 0.105%

instances where a review or consideration is being conducted or considered

No Comments

No Known Activations