INDEX

Explanations

instances of the word "Em" preceded by an adjective or noun term

oai_token-act-pair · gpt-3.5-turbo

the occurrences of the term "Em" indicating emotional context or emphasis

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

ASC

-0.69

 Kubrick

-0.68

 Elves

-0.68

 Encyclopedia

-0.66

 Icelandic

-0.65

 Iceland

-0.64

bed

-0.61

 knot

-0.60

 Asgard

-0.59

 novelty

-0.59

POSITIVE LOGITS

phasis

1.42

otional

1.40

otions

1.35

otion

1.28

igrant

1.26

brace

1.23

igrants

1.22

igration

1.21

perors

1.21

phas

1.19

Activations Density 0.026%

instances of the word "Em" preceded by an adjective or noun term

the occurrences of the term "Em" indicating emotional context or emphasis

No Comments

No Known Activations