INDEX

Explanations

negative events or outcomes in various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases or terms related to challenges and failures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

uana

-0.85

epad

-0.79

anguage

-0.74

iciency

-0.68

 translator

-0.66

keleton

-0.65

hirt

-0.63

vantage

-0.63

 wallpaper

-0.61

POSITIVE LOGITS

 victories

1.66

 incidents

1.60

 outings

1.59

 successes

1.57

 failures

1.56

 disasters

1.55

 collapses

1.53

 occasions

1.53

 tragedies

1.50

 defeats

1.49

Activations Density 0.858%

negative events or outcomes in various contexts

phrases or terms related to challenges and failures

No Comments

No Known Activations