INDEX

Explanations

names or terms that end in "ht" or "HT"

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atories

-0.68

ãĥ£

-0.67

clusive

-0.66

ãĤŃ

-0.63

ately

-0.62

fman

-0.61

hips

-0.60

yip

-0.60

cean

-0.59

ce

-0.59

POSITIVE LOGITS

ulhu

1.09

TPS

1.05

inen

0.94

icket

0.93

ronics

0.89

unia

0.87

hered

0.82

ras

0.81

ronic

0.81

opia

0.80

Activations Density 0.026%

names or terms that end in "ht" or "HT"

No Comments

No Known Activations