INDEX

Explanations

instances of discussions or controversies related to politics and policies

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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´ãĥ³

-0.80

alo

-0.72

ãĤ¦ãĤ¹

-0.68

alogy

-0.64

 afterlife

-0.63

opia

-0.63

vre

-0.63

 prev

-0.61

Widget

-0.61

Probably

-0.60

POSITIVE LOGITS

 however

0.77

when

0.71

 Newsweek

0.69

 rumors

0.68

 Attorney

0.67

 VICE

0.67

 tensions

0.67

 researchers

0.66

 Stephanie

0.65

 WikiLeaks

0.65

Activations Density 2.497%

instances of discussions or controversies related to politics and policies

No Comments

No Known Activations