INDEX

Explanations

references to the word "sheep"

oai_token-act-pair · gpt-3.5-turbo

references to sheep and lambs

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

AU

-0.73

nic

-0.72

 Alicia

-0.72

anc

-0.71

cial

-0.70

ondo

-0.69

Atl

-0.68

Dur

-0.68

 Extensions

-0.67

cade

-0.67

POSITIVE LOGITS

 sheep

3.93

 Sheep

2.53

 goats

1.92

 chickens

1.73

 cattle

1.65

 lamb

1.65

 goat

1.62

 wool

1.58

 livestock

1.58

 cows

1.52

Activations Density 0.011%

references to the word "sheep"

references to sheep and lambs

No Comments

No Known Activations

references to the word "sheep"

references to sheep and lambs

No Comments

No Known Activations