INDEX

Explanations

information about political viewpoints and policy positions

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 teasp

-0.91

OGR

-0.84

à¥

-0.77

quez

-0.74

photo

-0.74

oland

-0.73

à¤¾

-0.72

KK

-0.72

retty

-0.72

KO

-0.72

POSITIVE LOGITS

how

1.12

 whether

0.86

why

0.82

 them

0.81

 what

0.79

 calib

0.79

him

0.79

the

0.76

 ourselves

0.76

our

0.74

Activations Density 0.190%

information about political viewpoints and policy positions

No Comments

No Known Activations