INDEX

Explanations

specific mentions of the word "intention."

oai_token-act-pair · gpt-3.5-turbo

terms related to intention and evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 straw

-0.70

scl

-0.64

creen

-0.62

mind

-0.61

 tipping

-0.59

 Springer

-0.58

 smugglers

-0.57

jay

-0.56

cart

-0.56

 cart

-0.56

POSITIVE LOGITS

ally

1.93

ality

1.77

als

1.51

alities

1.51

arily

1.49

alist

1.47

ary

1.43

aries

1.36

nel

1.31

al

1.23

Activations Density 0.254%

specific mentions of the word "intention."

terms related to intention and evaluation

No Comments

No Known Activations