INDEX

Explanations

negative values or phrases indicating a decline

oai_token-act-pair · gpt-4o-mini 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

inant

-0.66

©¶æ

-0.65

 Primordial

-0.64

 opin

-0.62

ocus

-0.61

hiba

-0.61

 guiIcon

-0.58

 sentient

-0.58

intosh

-0.57

 intrins

-0.57

POSITIVE LOGITS

ember

1.13

December

1.02

 September

1.01

September

1.01

June

0.97

 August

0.96

 October

0.96

 December

0.95

 July

0.95

 June

0.94

Activations Density 0.078%

negative values or phrases indicating a decline

No Comments

No Known Activations