INDEX

Explanations

words related to technology or measurements

oai_token-act-pair · gpt-3.5-turbo

references to software testing or quality assurance terminology

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ItemTracker

-0.75

AMI

-0.70

 coefficient

-0.64

CVE

-0.62

 restraints

-0.61

eering

-0.61

HIP

-0.60

osate

-0.60

 Bind

-0.60

ressive

-0.60

POSITIVE LOGITS

terness

1.11

ques

1.05

ters

1.04

boro

0.96

aurus

0.93

pler

0.90

cil

0.88

tes

0.87

chet

0.87

ter

0.86

Activations Density 0.018%

words related to technology or measurements

references to software testing or quality assurance terminology

No Comments

No Known Activations