INDEX

Explanations

adverbs expressing agreement or correctness

oai_token-act-pair · gpt-3.5-turbo

words indicating justification or correctness

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fertility

-0.71

amen

-0.70

wives

-0.69

ession

-0.67

iments

-0.66

uese

-0.65

 Volunteers

-0.64

 formations

-0.63

 Occupations

-0.63

 portals

-0.62

POSITIVE LOGITS

ãĤ©

1.06

é¾į

0.92

eous

0.91

 deserved

0.86

 rightly

0.85

 rightfully

0.81

ãĥ£

0.76

assert

0.75

 deserves

0.74

 deserve

0.70

Activations Density 0.028%

adverbs expressing agreement or correctness

words indicating justification or correctness

No Comments

No Known Activations