INDEX

Explanations

the name "Narendra Modi"

oai_token-act-pair · gpt-3.5-turbo

mentions of political leaders, particularly Narendra Modi

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

 overhead

-0.67

 split

-0.61

 camp

-0.59

beck

-0.59

ps

-0.59

 crew

-0.59

loss

-0.58

POW

-0.58

 chim

-0.58

 correspondence

-0.58

POSITIVE LOGITS

 Narendra

3.59

 Modi

2.15

 Shinzo

1.88

 Recep

1.59

 Rahul

1.54

 Tayyip

1.53

 Gujar

1.53

 Gujarat

1.53

BJP

1.47

 Kejriwal

1.45

Activations Density 0.011%

the name "Narendra Modi"

mentions of political leaders, particularly Narendra Modi

No Comments

No Known Activations