INDEX

Explanations

prepositions and conjunctions indicating relationships or actions

oai_token-act-pair · gpt-3.5-turbo

words and phrases indicating conditions or exceptions in reasoning

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

sonian

-0.75

anger

-0.74

aban

-0.70

DAQ

-0.69

obin

-0.66

ouf

-0.63

anamo

-0.62

istani

-0.60

AMI

-0.60

bernatorial

-0.59

POSITIVE LOGITS

ĨĴ

0.64

 majority

0.62

jong

0.61

"{

0.60

Ħ¢

0.59

ideo

0.57

 spite

0.56

©¶æ

0.54

+(

0.54

 SOME

0.53

Activations Density 0.246%

prepositions and conjunctions indicating relationships or actions

words and phrases indicating conditions or exceptions in reasoning

No Comments

No Known Activations