INDEX

Explanations

titles or names with the word "Robert"

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to notable individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ashtra

-0.82

pour

-0.67

achus

-0.64

ibo

-0.63

Marie

-0.63

uable

-0.62

voy

-0.60

itiveness

-0.59

igans

-0.59

warm

-0.59

POSITIVE LOGITS

ħĭ

0.80

ĵĺ

0.75

¶æ

0.69

 Curve

0.64

 Reich

0.63

²

0.63

 Wilhelm

0.63

eston

0.63

 Carney

0.62

ĻĤ

0.62

Activations Density 0.109%

titles or names with the word "Robert"

proper nouns related to notable individuals

No Comments

No Known Activations