INDEX

Explanations

references to tigers

oai_token-act-pair · gpt-3.5-turbo

references to tigers

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

âĪ

-0.72

ecc

-0.69

 cann

-0.69

 epis

-0.67

CW

-0.65

 caps

-0.62

 activ

-0.60

 unden

-0.60

resp

-0.60

 recip

-0.59

POSITIVE LOGITS

 Tiger

4.09

 tiger

2.34

 Tigers

2.17

 tigers

1.93

 Leopard

1.72

 Jaguar

1.55

 Panther

1.55

iger

1.37

 Typhoon

1.35

 Rhino

1.34

Activations Density 0.026%

references to tigers

references to tigers

No Comments

No Known Activations