INDEX

Explanations

mentions of the word "cannibal."

oai_token-act-pair · gpt-3.5-turbo

terms related to cannibalism and associated concepts

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

-0.72

 Wink

-0.72

 sunscreen

-0.70

Weiss

-0.69

chest

-0.69

Rose

-0.68

 Rose

-0.67

 benz

-0.66

 Peach

-0.64

Berry

-0.64

POSITIVE LOGITS

ibal

3.81

 cannibal

2.13

 Hannibal

1.68

esar

1.16

abal

1.12

 butcher

1.06

atorial

1.04

 Quentin

1.02

fred

0.99

Rak

0.94

Activations Density 0.075%

mentions of the word "cannibal."

terms related to cannibalism and associated concepts

No Comments

No Known Activations