INDEX

Explanations

references to the word "Orange" in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Orange."

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

sonian

-0.90

stood

-0.87

Ö¼

-0.84

DIR

-0.84

enance

-0.78

ebin

-0.76

ted

-0.75

arnaev

-0.75

tle

-0.75

schild

-0.74

POSITIVE LOGITS

 Blossom

1.10

 Orange

1.00

 Peel

0.98

 Juice

0.97

 peel

0.94

vale

0.94

berry

0.89

 juice

0.84

Orange

0.82

 Crush

0.77

Activations Density 0.004%

references to the word "Orange" in various contexts

mentions of the word "Orange."

No Comments

No Known Activations