INDEX

Explanations

strong emotions or intense actions, especially related to negative events or information

oai_token-act-pair · gpt-3.5-turbo

words related to fear or being scared

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

 remembrance

-0.67

 shorth

-0.62

 contender

-0.62

 laure

-0.61

 Lauder

-0.60

 Lyme

-0.60

 tender

-0.58

 known

-0.58

 standby

-0.58

 Arbit

-0.58

POSITIVE LOGITS

ratch

1.59

enario

1.48

ulpt

1.44

reenshots

1.43

reenshot

1.41

issors

1.39

reens

1.36

arcity

1.32

enery

1.29

attered

1.29

Activations Density 0.016%

strong emotions or intense actions, especially related to negative events or information

words related to fear or being scared

No Comments

No Known Activations