INDEX

Explanations

mentions of individuals and their actions or statements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĤ¼

-0.65

 Enhance

-0.65

 Beaut

-0.64

 Dise

-0.63

ishable

-0.63

 Travels

-0.62

 Masquerade

-0.61

 Cutting

-0.61

 Patch

-0.61

MAP

-0.61

POSITIVE LOGITS

 replied

1.53

 answered

1.40

 reply

1.38

 replies

1.38

 responded

1.34

 answer

1.29

 hesitated

1.25

 answers

1.16

 answ

1.12

Answer

1.11

Activations Density 0.201%

mentions of individuals and their actions or statements

No Comments

No Known Activations