INDEX

Explanations

proper nouns or entities related to establishment or creation

oai_token-act-pair · gpt-3.5-turbo

terms related to origins and beginnings of entities or concepts

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

anton

-0.74

helm

-0.71

cer

-0.69

Ã°

-0.66

 herself

-0.66

 Mack

-0.65

lund

-0.64

lear

-0.63

inson

-0.62

ston

-0.61

POSITIVE LOGITS

 counterparts

0.92

ãĥķãĤ©

0.85

 counterpart

0.83

 holdings

0.79

 brethren

0.76

 footprint

0.74

 predecessors

0.71

 acronym

0.70

 woes

0.70

 contents

0.70

Activations Density 0.349%

proper nouns or entities related to establishment or creation

terms related to origins and beginnings of entities or concepts

No Comments

No Known Activations