INDEX

Explanations

mentions of law schools and related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to law and legal education

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

ħĭ

-0.74

ibilities

-0.69

 Philips

-0.68

awa

-0.66

attery

-0.65

 retina

-0.65

 Carbuncle

-0.64

avorite

-0.64

emonic

-0.64

oresc

-0.61

POSITIVE LOGITS

yer

1.50

suits

1.33

suit

1.33

yers

1.25

makers

1.10

 Enforcement

1.05

rence

1.03

maker

0.96

rences

0.90

lessness

0.90

Activations Density 0.033%

mentions of law schools and related terms

terms related to law and legal education

No Comments

No Known Activations