INDEX

Explanations

certain phrases that express a judgment or critique

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 respir

-0.65

 Secondly

-0.59

ABE

-0.55

 Regions

-0.54

SOURCE

-0.53

/#

-0.52

 resusc

-0.51

DOWN

-0.51

riger

-0.51

nih

-0.50

POSITIVE LOGITS

pires

1.12

pired

1.11

piring

0.93

 follows

0.93

pire

0.89

ynchron

0.89

eline

0.85

leep

0.85

ylum

0.85

piration

0.84

Activations Density 0.067%

certain phrases that express a judgment or critique

No Comments

No Known Activations