INDEX

Explanations

terms related to uncertainty, especially high levels of uncertainty

oai_token-act-pair · gpt-3.5-turbo

references to uncertainty 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

utsche

-0.84

ongyang

-0.84

itor

-0.82

ovie

-0.81

baugh

-0.80

artist

-0.77

ONY

-0.77

ingo

-0.70

Interstitial

-0.70

ERSON

-0.68

POSITIVE LOGITS

 surrounds

1.06

 surrounding

0.99

 arising

0.94

worthiness

0.87

 arises

0.85

 uncertainty

0.81

 shroud

0.79

lessly

0.78

 concerning

0.78

 swirling

0.76

Activations Density 0.055%

terms related to uncertainty, especially high levels of uncertainty

references to uncertainty in various contexts

No Comments

No Known Activations