INDEX

Explanations

expressions related to extension or expansion

oai_token-act-pair · gpt-3.5-turbo

references to support, expansion, or extension of various concepts or entities

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

bish

-0.64

EStream

-0.63

isky

-0.62

utan

-0.60

Ì

-0.60

apest

-0.60

kov

-0.60

PASS

-0.58

chem

-0.57

berg

-0.57

POSITIVE LOGITS

 bounds

0.95

izons

0.86

 boundaries

0.79

 scope

0.77

 distances

0.75

 outward

0.75

 horizont

0.74

 horizontally

0.74

 breadth

0.74

 diameter

0.74

Activations Density 0.160%

expressions related to extension or expansion

references to support, expansion, or extension of various concepts or entities

No Comments

No Known Activations