INDEX

Explanations

military and governmental operation related phrases

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

Negative Logits

Generic

-0.69

Pros

-0.69

Chair

-0.64

MD

-0.62

ILE

-0.61

ahime

-0.61

TABLE

-0.60

atography

-0.59

ilic

-0.58

lance

-0.57

POSITIVE LOGITS

 possible

0.90

yip

0.88

pload

0.87

thirds

0.81

 dozen

0.80

 eighty

0.78

 practicable

0.76

 half

0.72

dozen

0.70

0.69

Activations Density 0.032%

military and governmental operation related phrases

No Known Activations