INDEX

Explanations

phrases related to ranking, comparison, or superlatives

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

izon

-0.85

vous

-0.75

neys

-0.70

pty

-0.70

md

-0.70

kay

-0.66

prototype

-0.66

skirts

-0.66

lins

-0.64

neg

-0.64

POSITIVE LOGITS

 bidder

1.15

 rated

0.98

 level

0.93

 earners

0.93

 priority

0.92

 concentrations

0.91

 levels

0.90

 tier

0.90

 stakes

0.90

 priced

0.89

Activations Density 0.023%

phrases related to ranking, comparison, or superlatives

No Comments

No Known Activations