INDEX

Explanations

phrases related to qualities, trends, and behavior without a specific focus on a single topic

oai_token-act-pair · gpt-3.5-turbo

phrases related to the reputation or notoriety of various 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

viz

-0.61

 obser

-0.58

?".

-0.56

 behav

-0.56

would

-0.55

Is

-0.54

 thereof

-0.54

actual

-0.53

tnc

-0.53

OTO

-0.53

POSITIVE LOGITS

 Dempsey

0.58

 decentral

0.55

 saddened

0.55

 hugely

0.53

 buoy

0.53

 decentralized

0.52

 synonymous

0.52

steamapps

0.51

ablished

0.51

 premiere

0.51

Activations Density 0.888%

phrases related to qualities, trends, and behavior without a specific focus on a single topic

phrases related to the reputation or notoriety of various entities

No Comments

No Known Activations