INDEX

Explanations

phrases indicating a reversal or reconsideration

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

ievers

-0.59

quartered

-0.58

styles

-0.58

aneers

-0.57

aukee

-0.57

ãĥ³ãĤ¸

-0.56

 contracted

-0.56

riors

-0.56

usha

-0.55

apons

-0.55

POSITIVE LOGITS

worldly

0.71

IRD

0.68

Ħ¢

0.66

 happening

0.65

OULD

0.64

lander

0.64

GD

0.60

mentioned

0.60

ALE

0.59

creen

0.58

Activations Density 0.208%

phrases indicating a reversal or reconsideration

No Comments

No Known Activations