INDEX

Explanations

phrases related to negative outcomes or criticism

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

FORMATION

-0.77

FIG

-0.73

shapeshifter

-0.70

uzzle

-0.70

SPONSORED

-0.68

PLA

-0.67

AY

-0.67

çĶ°

-0.67

iewicz

-0.65

taboola

-0.65

POSITIVE LOGITS

 altogether

1.47

 inhib

1.04

 entirely

1.04

 prematurely

0.90

 virginity

0.88

 restraints

0.83

 pesky

0.83

 shack

0.81

 reliance

0.80

 reins

0.78

Activations Density 2.105%

phrases related to negative outcomes or criticism

No Comments

No Known Activations