INDEX

Explanations

contradictory statements or phrases that express contrasting ideas

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

onte

-0.81

ife

-0.77

ENG

-0.72

ãĥĺ

-0.72

æ°

-0.71

locked

-0.70

lled

-0.69

ãĤ¼ãĤ¦ãĤ¹

-0.67

çĶ°

-0.64

enter

-0.64

POSITIVE LOGITS

 nonetheless

1.12

 nevertheless

1.09

 suffice

1.00

 insofar

0.87

 certainly

0.82

 judging

0.79

 shouldn

0.76

 surely

0.74

 facts

0.73

 pointing

0.73

Activations Density 0.273%

contradictory statements or phrases that express contrasting ideas

No Comments

No Known Activations