INDEX

Explanations

words expressing strong emotions or attitudes

oai_token-act-pair · gpt-3.5-turbo

expressions of intense feelings or emotions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

adelphia

-0.81

sburgh

-0.77

olor

-0.73

ourses

-0.71

onis

-0.69

ses

-0.67

ensis

-0.66

osal

-0.65

heid

-0.65

 Extensions

-0.63

POSITIVE LOGITS

 important

0.93

 informative

0.91

 rare

0.89

 difficult

0.89

 interesting

0.88

 unlikely

0.86

 exciting

0.85

 transparent

0.84

 useful

0.82

 busy

0.82

Activations Density 0.065%

words expressing strong emotions or attitudes

expressions of intense feelings or emotions

No Comments

No Known Activations