INDEX

Explanations

names or terms related to a person named "Ab"

oai_token-act-pair · gpt-3.5-turbo

mentions of individuals or entities referred to as "Ab."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 subpoen

-0.74

 lemon

-0.71

 Lemon

-0.70

 Maple

-0.68

 mint

-0.67

 grou

-0.67

 chill

-0.67

gre

-0.67

 standalone

-0.66

 pole

-0.66

POSITIVE LOGITS

Ab

3.83

Ab

2.02

Ali

1.43

AB

1.38

 Abram

1.35

Abs

1.35

Ag

1.30

ab

1.28

Abd

1.25

Ak

1.23

Activations Density 0.011%

names or terms related to a person named "Ab"

mentions of individuals or entities referred to as "Ab."

No Comments

No Known Activations