INDEX

Explanations

phrases related to providing explanations or reasons

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate explanations or justifications

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

ography

-0.77

emies

-0.76

ctors

-0.74

nets

-0.74

heit

-0.71

Dialogue

-0.71

jab

-0.69

dayName

-0.69

nown

-0.68

ograp

-0.67

POSITIVE LOGITS

why

1.53

why

1.13

WHY

0.99

 discrepancies

0.97

 variance

0.91

 reluctance

0.89

 inconsistencies

0.84

 discrep

0.82

how

0.80

 disparities

0.79

Activations Density 0.126%

phrases related to providing explanations or reasons

phrases that indicate explanations or justifications

No Comments

No Known Activations