INDEX

Explanations

phrases related to legal rulings or decisions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ruled."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Load

-0.71

 packages

-0.70

 load

-0.70

 Spray

-0.68

 package

-0.68

lift

-0.64

rawdownloadcloneembedreportprint

-0.63

 Focus

-0.61

 pickup

-0.61

Activ

-0.61

POSITIVE LOGITS

 ruled

3.83

 rule

1.74

 governed

1.72

 ruling

1.70

rul

1.67

rule

1.52

Rule

1.39

uled

1.39

 upheld

1.36

 presided

1.34

Activations Density 0.015%

phrases related to legal rulings or decisions

occurrences of the word "ruled."

No Comments

No Known Activations