INDEX

Explanations

phrases related to definitions and legal interpretations

oai_token-act-pair · gpt-3.5-turbo

references to definitions and concepts related to various topics and issues

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

 heels

-0.78

ersen

-0.77

 outweigh

-0.68

heny

-0.68

olulu

-0.67

umbn

-0.65

jri

-0.63

athi

-0.62

ilant

-0.62

ilyn

-0.61

POSITIVE LOGITS

 thresholds

0.72

 boundaries

0.71

 Species

0.69

verages

0.68

Spec

0.67

Characters

0.66

 quantity

0.66

 parameters

0.65

 Individual

0.64

à¼

0.62

Activations Density 0.149%

phrases related to definitions and legal interpretations

references to definitions and concepts related to various topics and issues

No Comments

No Known Activations