INDEX

Explanations

actions involving violent or controversial incidents, including physical altercations and political or military operations

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

\"

-0.69

 deadlines

-0.64

 workflow

-0.63

 delivering

-0.61

doors

-0.60

 sabot

-0.60

 releasing

-0.60

Impro

-0.59

 Practices

-0.59

 collabor

-0.58

POSITIVE LOGITS

 wonder

1.08

 speculate

1.02

 conjecture

0.97

 wondered

0.96

 imagine

0.95

 extrap

0.92

 remem

0.91

athom

0.91

 recall

0.90

 shudder

0.90

Activations Density 2.800%

actions involving violent or controversial incidents, including physical altercations and political or military operations

No Comments

No Known Activations