INDEX

Explanations

language that indicates substantial or significant changes or impacts

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

ifully

-0.87

elsen

-0.81

sburgh

-0.78

afety

-0.76

tein

-0.75

ventus

-0.72

icip

-0.72

 Memories

-0.70

atu

-0.69

arest

-0.69

POSITIVE LOGITS

 impacted

0.97

 impacting

0.95

 effected

0.91

 reduces

0.88

 reduce

0.88

 outper

0.88

 differ

0.86

 reduced

0.86

 underest

0.85

 altered

0.85

Activations Density 0.018%

language that indicates substantial or significant changes or impacts

No Comments

No Known Activations