INDEX

Explanations

phrases related to a specific organization or entity

oai_token-act-pair · gpt-3.5-turbo

references to political assemblies or legislative bodies

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

 policies

-0.63

everything

-0.61

font

-0.59

 Giants

-0.58

Poe

-0.58

NO

-0.57

 everything

-0.56

 timestamp

-0.56

gif

-0.55

 Everything

-0.54

POSITIVE LOGITS

 Sabha

5.34

 Pradesh

1.23

sembly

1.08

 Lumpur

1.04

 Nadu

1.02

 Tripoli

0.96

aughtered

0.91

 Lanka

0.90

ashtra

0.89

 Krish

0.88

Activations Density 0.015%

phrases related to a specific organization or entity

references to political assemblies or legislative bodies

No Comments

No Known Activations