INDEX

Explanations

direct or implied commands

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

productive

-0.68

Kop

-0.65

////////////////////////////////

-0.65

 Feld

-0.64

ults

-0.61

 Appears

-0.61

 Survivors

-0.61

quickShipAvailable

-0.59

 genesis

-0.57

 Matters

-0.57

POSITIVE LOGITS

 join

1.09

 clarify

1.09

 intervene

1.08

 participate

1.06

 leave

1.05

 accompany

1.05

 submit

1.03

 nominate

1.03

 give

1.02

 remove

1.01

Activations Density 0.087%

direct or implied commands

No Comments

No Known Activations