INDEX

Explanations

imperatives or questions requesting action or information

oai_token-act-pair · gpt-3.5-turbo

interrogative phrases and elements indicating state or existence

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

 alot

-0.78

 doesnt

-0.74

nor

-0.70

erent

-0.68

 didnt

-0.66

 dont

-0.65

other

-0.62

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.60

DragonMagazine

-0.59

 independ

-0.59

POSITIVE LOGITS

—"

0.88

0.80

—

0.79

—

0.78

0.73

.)

0.73

0.72

0.71

--

0.69

,—

0.67

Activations Density 0.872%

imperatives or questions requesting action or information

interrogative phrases and elements indicating state or existence

No Comments

No Known Activations