INDEX

Explanations

phrases related to assistance, contribution, and helping

oai_token-act-pair · gpt-3.5-turbo

phrases that imply contributing or helping actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Procedure

-0.66

Statement

-0.63

 behaves

-0.58

 awaits

-0.57

llo

-0.57

 Phant

-0.57

 Response

-0.56

 faced

-0.56

Jol

-0.56

 lashes

-0.56

POSITIVE LOGITS

 perpetuate

1.03

 alleviate

1.00

 strengthen

0.98

 refine

0.96

 reinforce

0.93

ãĥİ

0.92

 improve

0.91

 enhance

0.90

 increase

0.89

 dispel

0.88

Activations Density 0.103%

phrases related to assistance, contribution, and helping

phrases that imply contributing or helping actions

No Comments

No Known Activations