INDEX

Explanations

positive or negative evaluations

oai_token-act-pair · gpt-3.5-turbo

phrases expressing difficulty or negative assessments

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

isd

-0.71

pert

-0.68

release

-0.68

ternal

-0.65

last

-0.64

cellence

-0.63

utic

-0.62

Offline

-0.62

clamation

-0.62

ruption

-0.61

POSITIVE LOGITS

 anymore

0.89

nor

0.87

MpServer

0.69

oshenko

0.67

leans

0.66

 slicing

0.64

 elbows

0.62

 whatsoever

0.60

NJ

0.60

DEM

0.60

Activations Density 0.082%

positive or negative evaluations

phrases expressing difficulty or negative assessments

No Comments

No Known Activations