INDEX

Explanations

mentions of the word "humans" in the text

oai_token-act-pair · gpt-3.5-turbo

mentions of humans

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

 Crom

-0.71

ounter

-0.66

sbm

-0.64

pton

-0.63

tie

-0.63

 exclusive

-0.62

orama

-0.61

 Style

-0.61

 magazine

-0.61

pin

-0.61

POSITIVE LOGITS

 humans

3.59

 Humans

2.81

humans

2.52

 human

2.11

 humankind

2.10

 mortals

2.00

 mammals

1.91

 primates

1.84

human

1.83

 humanity

1.82

Activations Density 0.020%

mentions of the word "humans" in the text

mentions of humans

No Comments

No Known Activations