INDEX

Explanations

phrases expressing comparisons or preferences

oai_token-act-pair · gpt-3.5-turbo

phrases indicating improvement or betterment

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

igham

-0.82

quin

-0.68

atan

-0.61

 unpop

-0.59

 scissors

-0.56

mod

-0.56

Cont

-0.55

 Command

-0.55

itars

-0.55

ADD

-0.55

POSITIVE LOGITS

than

1.17

 than

1.15

 Than

1.05

 chances

0.77

 outcomes

0.69

aroo

0.66

 spelling

0.62

 Recon

0.62

oldemort

0.61

iate

0.61

Activations Density 0.155%

phrases expressing comparisons or preferences

phrases indicating improvement or betterment

No Comments

No Known Activations