INDEX

Explanations

positive and encouraging expressions, wishes, and sentiments through words like "Good luck," "Thanks for reading," and "Hope you enjoy."

oai_token-act-pair · gpt-3.5-turbo

expressions of good luck and encouragement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.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

 helicop

-0.79

atton

-0.66

 blamed

-0.62

arthed

-0.62

uries

-0.61

ĨĴ

-0.61

 unilaterally

-0.60

 graft

-0.60

 Spani

-0.59

roup

-0.58

POSITIVE LOGITS

sir

0.87

:)

0.75

 Congratulations

0.73

0.71

 please

0.70

 LORD

0.68

reetings

0.68

 Thank

0.68

 Merry

0.67

 ladies

0.67

Activations Density 0.450%

positive and encouraging expressions, wishes, and sentiments through words like "Good luck," "Thanks for reading," and "Hope you enjoy."

expressions of good luck and encouragement

No Comments

No Known Activations