INDEX

Explanations

phrases that indicate dependency or variability

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate conditions or factors influencing outcomes

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

©¶æ¥µ

-0.85

lyn

-0.77

ãĥĨ

-0.75

jah

-0.73

together

-0.73

Shield

-0.72

Li

-0.71

utsch

-0.70

µ

-0.70

Hall

-0.67

POSITIVE LOGITS

 circumstances

1.14

 circumstance

1.05

 whether

0.99

how

0.96

 factors

0.91

 severity

0.86

 prevailing

0.82

 behalf

0.81

 weather

0.81

 occasion

0.81

Activations Density 0.068%

phrases that indicate dependency or variability

phrases that indicate conditions or factors influencing outcomes

No Comments

No Known Activations