INDEX

Explanations

adjectives and nouns that describe the characteristics and behaviors of objects, concepts or individuals

oai_token-act-pair · gpt-3.5-turbo

phrases that highlight commonly accepted beliefs or generalizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

inventoryQuantity

-0.76

skirts

-0.75

atures

-0.73

ctors

-0.72

SI

-0.69

ntil

-0.69

ciating

-0.68

might

-0.67

ĺ

-0.66

Stars

-0.66

POSITIVE LOGITS

 confined

0.85

 excluded

0.82

 mistaken

0.80

 regarded

0.80

 considered

0.76

 connected

0.75

 indistinguishable

0.73

 tied

0.73

 compelled

0.73

 eliminated

0.72

Activations Density 0.203%

adjectives and nouns that describe the characteristics and behaviors of objects, concepts or individuals

phrases that highlight commonly accepted beliefs or generalizations

No Comments

No Known Activations