INDEX

Explanations

references to measurements in yards

oai_token-act-pair · gpt-3.5-turbo

instances of the word "yard" in various contexts

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

ipper

-0.77

omal

-0.72

isen

-0.72

uten

-0.70

jur

-0.68

 coma

-0.67

 inval

-0.67

ivia

-0.66

emic

-0.65

emia

-0.65

POSITIVE LOGITS

yard

1.35

yards

1.16

birds

0.92

 yard

0.84

acre

0.84

arious

0.80

wagon

0.80

 Yard

0.76

igans

0.76

cakes

0.74

Activations Density 0.004%

references to measurements in yards

instances of the word "yard" in various contexts

No Comments

No Known Activations