INDEX

Explanations

professionals and their respective roles or activities

oai_token-act-pair · gpt-3.5-turbo

actions related to individuals' occupations and their roles in specific contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 deterrent

-0.74

vity

-0.73

 moot

-0.70

 unbeliev

-0.69

rary

-0.68

ricks

-0.64

 cancellation

-0.64

 exaggeration

-0.62

 Peoples

-0.62

unless

-0.61

POSITIVE LOGITS

ufact

0.68

ifa

0.68

 aged

0.64

sup

0.63

 slaughtered

0.62

 retiring

0.62

pires

0.61

jong

0.61

 sacrificed

0.61

 eaten

0.61

Activations Density 1.321%

professionals and their respective roles or activities

actions related to individuals' occupations and their roles in specific contexts

No Comments

No Known Activations