INDEX

Explanations

mentions of leadership, particularly in a critical or influential context

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

 loopholes

-0.65

 mercy

-0.64

aves

-0.63

 symmetry

-0.58

ALLY

-0.57

entric

-0.57

irlf

-0.57

 Vaughn

-0.56

fav

-0.55

rehens

-0.55

POSITIVE LOGITS

better

0.96

gers

0.96

hunt

0.86

ges

0.81

ership

0.79

quartered

0.78

wig

0.75

gling

0.74

-+

0.72

gew

0.72

Activations Density 1.435%

mentions of leadership, particularly in a critical or influential context

No Comments

No Known Activations