INDEX

Explanations

phrases related to events, political actions, and organized groups

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

ear

-0.69

 unchecked

-0.64

 supremacy

-0.62

enberg

-0.60

 Patriarch

-0.59

perm

-0.59

 mamm

-0.58

Ĥİ

-0.57

thur

-0.57

cale

-0.57

POSITIVE LOGITS

hips

1.00

imental

0.83

 therein

0.81

lees

0.78

iments

0.75

iltr

0.74

iltration

0.74

cius

0.74

enza

0.73

lished

0.73

Activations Density 5.312%

phrases related to events, political actions, and organized groups

No Comments

No Known Activations