INDEX

Explanations

instances of specific individuals or entities being discussed in varying contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

 unlucky

-0.71

 insignificant

-0.70

 irresistible

-0.69

 nuts

-0.68

 enslaved

-0.67

 parasitic

-0.66

 parity

-0.65

 delinquent

-0.65

 hiber

-0.65

 asteroids

-0.65

POSITIVE LOGITS

 Photo

1.24

 Photograph

1.20

 PHOTO

1.19

 Courtesy

1.16

jpg

1.15

 Photographer

1.11

Photo

1.06

 REUTERS

1.05

 Image

1.05

 Picture

1.04

Activations Density 0.124%

instances of specific individuals or entities being discussed in varying contexts

No Comments

No Known Activations