INDEX

Explanations

pronouns and verbs related to personal interactions and conversations

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 Pastebin

-0.67

town

-0.67

milo

-0.65

ĸļ

-0.64

 Coalition

-0.62

 Sutherland

-0.61

 Census

-0.61

 Said

-0.57

 Houses

-0.56

 congress

-0.56

POSITIVE LOGITS

 irresist

0.91

 closer

0.88

 stumble

0.77

 reconsider

0.75

 susceptible

0.75

awa

0.73

 immune

0.72

oser

0.72

 psychologically

0.72

 rethink

0.72

Activations Density 1.173%

pronouns and verbs related to personal interactions and conversations

No Comments

No Known Activations