INDEX

Explanations

positive sentiments towards certain actions or behaviors

oai_token-act-pair · gpt-3.5-turbo

words indicating acceptance and appreciation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

orio

-0.62

 Milky

-0.59

guyen

-0.58

ouf

-0.54

 coordinates

-0.53

ierre

-0.53

 occupancy

-0.53

 Delay

-0.52

lua

-0.52

iencies

-0.51

POSITIVE LOGITS

by

1.26

By

0.93

 universally

0.92

by

0.91

 widely

0.91

 enthusiastically

0.90

 everywhere

0.90

 amongst

0.89

 internationally

0.89

 nationally

0.88

Activations Density 0.219%

positive sentiments towards certain actions or behaviors

words indicating acceptance and appreciation

No Comments

No Known Activations