INDEX

Explanations

single letters or abbreviations related to various topics

oai_token-act-pair · gpt-3.5-turbo

the letter "E" and its different occurrences throughout the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ifier

-0.86

ifiers

-0.74

MSM

-0.64

ĸļ

-0.63

ifiable

-0.63

anova

-0.63

lust

-0.61

frames

-0.61

 pains

-0.60

KM

-0.59

POSITIVE LOGITS

lements

1.15

LECT

1.09

instein

0.99

lected

0.97

leanor

0.94

agles

0.93

ITH

0.93

commerce

0.92

ASY

0.92

ighty

0.91

Activations Density 0.035%

single letters or abbreviations related to various topics

the letter "E" and its different occurrences throughout the text

No Comments

No Known Activations