INDEX

Explanations

verbs related to storytelling

oai_token-act-pair · gpt-3.5-turbo

references to manipulation or fabrication of narratives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Torrent

-0.72

sterdam

-0.70

trak

-0.69

 Collector

-0.66

 Dynamo

-0.65

stown

-0.64

 collector

-0.63

nin

-0.62

 Writ

-0.61

 Penguin

-0.60

POSITIVE LOGITS

 excuses

1.10

 excuse

1.09

 blaming

1.07

 justification

1.07

 assertion

0.96

 explanations

0.94

 accusing

0.94

 assertions

0.93

 blame

0.93

 exagger

0.91

Activations Density 0.750%

verbs related to storytelling

references to manipulation or fabrication of narratives

No Comments

No Known Activations