INDEX

Explanations

terms related to historical events or subjects

oai_token-act-pair · gpt-3.5-turbo

references to the concept of history

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

igans

-0.70

vae

-0.65

leased

-0.64

enger

-0.63

aton

-0.63

orc

-0.63

spir

-0.62

activate

-0.61

raised

-0.60

legate

-0.60

POSITIVE LOGITS

 textbooks

1.23

 buffs

1.19

 books

1.12

 lesson

1.07

 documentaries

0.92

 repeats

0.87

 professor

0.86

 textbook

0.86

 repeating

0.84

 lessons

0.84

Activations Density 0.069%

terms related to historical events or subjects

references to the concept of history

No Comments

No Known Activations