INDEX

Explanations

texts related to academic writing and formal language, potentially including math symbols

oai_token-act-pair · gpt-3.5-turbo

references to personal experiences and emotional reflections

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

',"

-0.71

imentary

-0.70

"],

-0.69

kefeller

-0.68

];

-0.66

igent

-0.65

 Newark

-0.63

];

-0.62

odge

-0.62

ogether

-0.61

POSITIVE LOGITS

âĢ

1.23

 anime

1.15

 Elsa

1.14

 Overwatch

1.10

 fandom

1.10

 Blizz

1.08

 GamerGate

1.05

 reddit

1.02

 Gamergate

1.01

ãĢİ

1.01

Activations Density 0.905%

texts related to academic writing and formal language, potentially including math symbols

references to personal experiences and emotional reflections

No Comments

No Known Activations