INDEX

Explanations

technical or scientific terms related to materials or compounds

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

 suspic

-0.78

uate

-0.67

 Reconstruction

-0.61

uated

-0.60

 releg

-0.60

 inhibition

-0.58

 distracted

-0.58

 depress

-0.58

 chang

-0.57

senal

-0.57

POSITIVE LOGITS

riter

1.25

ritten

1.17

estern

1.12

een

1.11

sburg

1.06

ITNESS

1.04

atcher

1.03

esome

1.03

olves

1.02

alker

1.01

Activations Density 0.021%

technical or scientific terms related to materials or compounds

No Comments

No Known Activations