INDEX

Explanations

statements made by individuals

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ©

-0.84

CLA

-0.81

ictionary

-0.79

âĨ

-0.78

Plus

-0.77

ãĤ¦ãĤ¹

-0.77

umblr

-0.75

Bottom

-0.75

brow

-0.75

rette

-0.74

POSITIVE LOGITS

 Armenia

1.17

 Azerbaijan

1.16

 Ankara

1.13

 militants

1.12

 Daesh

1.12

 Nato

1.10

 Azerbai

1.09

 Houth

1.09

 Hezbollah

1.08

 Turkey

1.07

Activations Density 0.325%

statements made by individuals

No Comments

No Known Activations