INDEX

Explanations

phrases related to prevention or avoidance

oai_token-act-pair · gpt-3.5-turbo

phrases indicating an intention to prevent or avoid negative outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ĸļ

-0.66

Tracker

-0.66

DEM

-0.64

Loc

-0.64

ALLY

-0.63

Services

-0.63

 listened

-0.63

Pak

-0.62

 largeDownload

-0.62

 ported

-0.61

POSITIVE LOGITS

 pitfalls

1.63

 temptation

1.05

 mistakes

1.04

 distractions

1.02

 embarrassment

1.02

 altogether

1.01

 hazards

1.00

 needless

0.98

 tragedies

0.97

 unnecessary

0.96

Activations Density 0.289%

phrases related to prevention or avoidance

phrases indicating an intention to prevent or avoid negative outcomes

No Comments

No Known Activations