INDEX

Explanations

references to units of weight, particularly the word "pound"

oai_token-act-pair · gpt-3.5-turbo

references to weight measurements

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

etheus

-0.77

OURCE

-0.77

PsyNetMessage

-0.75

 WARN

-0.71

cknow

-0.69

inav

-0.68

ILLE

-0.66

vironment

-0.65

mberg

-0.65

IRED

-0.65

POSITIVE LOGITS

 sterling

1.19

 Pound

1.13

cake

1.04

 pound

0.93

nuts

0.92

 Sterling

0.90

er

0.86

pound

0.85

ers

0.83

ings

0.80

Activations Density 0.014%

references to units of weight, particularly the word "pound"

references to weight measurements

No Comments

No Known Activations