INDEX

Explanations

mentions of the word "monkey"

oai_token-act-pair · gpt-3.5-turbo

mentions of monkeys or primates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

RAW

-0.91

Pg

-0.80

Solid

-0.80

soType

-0.77

âĢ¢âĢ¢âĢ¢âĢ¢

-0.76

Madison

-0.75

ãĥ´

-0.75

Avg

-0.75

JUST

-0.74

sburgh

-0.74

POSITIVE LOGITS

 monkey

1.15

 monkeys

1.14

oleon

0.98

 wrench

0.87

pox

0.87

 apes

0.85

 gorilla

0.81

esses

0.80

 bitten

0.80

patch

0.80

Activations Density 0.012%

mentions of the word "monkey"

mentions of monkeys or primates

No Comments

No Known Activations