INDEX

Explanations

legal and policy-related terms and concepts

oai_token-act-pair · gpt-3.5-turbo

references to legal or regulatory frameworks surrounding licenses and permits

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

stories

-0.75

Insp

-0.74

orge

-0.73

ULTS

-0.71

sbm

-0.69

uden

-0.69

udi

-0.67

rir

-0.66

akeru

-0.66

 NEWS

-0.65

POSITIVE LOGITS

 requirement

1.60

 exemption

1.52

 prohibition

1.37

 provision

1.37

 presumption

1.37

 threshold

1.35

 exception

1.33

 limitation

1.33

 clause

1.32

 penalty

1.32

Activations Density 0.324%

legal and policy-related terms and concepts

references to legal or regulatory frameworks surrounding licenses and permits

No Comments

No Known Activations