INDEX

Explanations

political references and statements by politicians

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

ãĥ¥

-0.68

worldly

-0.67

Russ

-0.65

zn

-0.63

Interested

-0.61

ARCH

-0.61

ARY

-0.59

oir

-0.58

âĶĢâĶĢâĶĢâĶĢ

-0.57

',

-0.57

POSITIVE LOGITS

 withdrew

0.82

 joins

0.82

 denies

0.80

was

0.74

 wrote

0.74

 appears

0.74

 became

0.74

 declined

0.73

 teaches

0.73

 opposes

0.72

Activations Density 0.943%

political references and statements by politicians

No Comments

No Known Activations