INDEX

Explanations

mentions of objects related to knives

oai_token-act-pair · gpt-3.5-turbo

mentions of knives and their use in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

isol

-0.75

ourced

-0.73

 Beckham

-0.73

bos

-0.73

Newsletter

-0.71

Demand

-0.68

leep

-0.68

mberg

-0.67

alse

-0.67

 Petro

-0.65

POSITIVE LOGITS

 knife

1.37

knife

1.35

 knives

1.33

 blade

1.12

 Knife

1.12

 blades

1.05

 scissors

0.91

powder

0.89

 dagger

0.87

 stabbed

0.86

Activations Density 0.010%

mentions of objects related to knives

mentions of knives and their use in various contexts

No Comments

No Known Activations