INDEX

Explanations

references to governmental or administrative accountability and actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

conn

-0.77

rolled

-0.76

eg

-0.76

tons

-0.76

furt

-0.76

gow

-0.75

sted

-0.74

seys

-0.74

utor

-0.73

ican

-0.72

POSITIVE LOGITS

 STORY

0.88

 PAGE

0.84

 DEFENSE

0.82

 NEWS

0.82

 NETWORK

0.80

OF

0.80

 QUEST

0.77

 EDITION

0.77

FAC

0.77

 INFORMATION

0.75

Activations Density 0.023%

references to governmental or administrative accountability and actions

No Comments

No Known Activations