INDEX

Explanations

words related to rules and regulations

oai_token-act-pair · gpt-3.5-turbo

terms related to rules and regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 focused

-0.73

 intensive

-0.70

 Harris

-0.68

 serial

-0.64

 engaged

-0.64

 Updates

-0.63

 dedicated

-0.61

 prof

-0.61

av

-0.60

 exposed

-0.59

POSITIVE LOGITS

rule

4.79

rules

1.91

Rule

1.49

rul

1.26

rite

1.16

rites

1.14

definition

1.08

Rules

1.08

ride

1.06

 rule

1.06

Activations Density 0.026%

words related to rules and regulations

terms related to rules and regulations

No Comments

No Known Activations